An approach towards C12 oxo analogues of the side chain of pumiliotoxin B/allopumiliotoxin 339A and B

Article abstract of DOI:10.1016/S0040-4039(02)01071-7

A route towards synthesis of analogues of the pumilotoxin/allopumilotoxin side-chain is described. The C15,C16 diol was introduced by asymmetric dihydroxylation using AD-mix β of C10,C17 eneyneone intermediate 14, or of C13,C17 precursor 17, or by using a chiron-based route from 24. The trisubstituted alkene functionality was established using thioaryl conjugate addition to yneones 16 and 27, followed by a copper-catalyzed stereoretentive reaction with methylmagnesium bromide. The approach enables access to C12 oxo systems and offers an approach towards new C14 analogues.

Full text of DOI:10.1016/S0040-4039(02)01071-7

TETRAHEDRON
LETTERS
Pergamon
Tetrahedron Letters 43 (2002) 5415–5418
An approach towards C12 oxo analogues of the side chain of
pumiliotoxin B/allopumiliotoxin 339A and B†
John M. Gardiner,* Philip E. Giles and Mar´ıa Luz Mart´ın Mart´ın‡
Department of Chemistry, University of Manchester Institute of Science and Technology, PO Box 88,
Manchester M60 1QD, UK
Received 12 April 2002; revised 10 May 2002; accepted 7 June 2002
Abstract—A route towards the synthesis of analogues of pumiliotoxin and allopumiliotoxin side-chain is described. The C15,C16
diol was introduced by asymmetric dihydroxylation using AD-mix b of C10,C17 enynone intermediate 14, or of C13,C17
precursor 17, or by using a chiron-based route from 24. The trisubstituted alkene functionality was established using arylthio
conjugate addition to ynones 16 and 27, followed by a copper-catalyzed stereoretentive reaction with methylmagnesium bromide.
The approach enables access to C12 oxo systems and offers an approach towards new C14 analogues. © 2002 Elsevier Science
Ltd. All rights reserved.
The venomous skin secretions of Dendrobatid frogs
have proven a rich source of structurally complex
bioactive alkaloids (several hundred), with a wide vari-
ety of potent biological activities,2 including potent
neurotoxins such as histrionicotoxins, batrachotoxin
and gephrotoxin. Pumiliotoxin and the closely related
rarer allopumiliotoxin alkaloids were first isolated over
30 years ago from this source.3 The architecture of the
structurally unusual pumiliotoxins A and B, and the
allopumiliotoxins (sharing the basic heterocyclic core of
pumiliotoxins A and B, differing only in possessing a
7-OH group) was established some years later (Fig. 1).4
Since then there has been much interest in these com-
pounds, spurred by their dramatic activity on cardiac
functions. The class of compounds is particularly inter-
esting, as some other members, differing only in side-
chain functionality or hydroxyl group protection, are
mild cardiodepressants.5 The cardiac activities dis-
played by pumiliotoxins, and the effects of changes in
side-chain functionality, indicates the importance of the
side chain to the mechanism of action of different
congeners. The mechanism of action of these agents
involves binding to voltage dependent sodium channels
and thereby modulating sodium flux through these
channels.5d,6
The allopumiliotoxins and pumiliotoxins have been the
subject of a number of synthetic efforts since the early
1980s. Total syntheses of the allopumiliotoxins 339A 2
and 339B 3 have been described by Trost,7 Overman8
and Kibayashi,9 and of pumiliotoxin B 1,10 along with
a number of approaches to the ring system. The Over-
man and Kibayashi syntheses proceed via an acetylenic
precursor of the side chain, or through a C10 aldehyde
precursor, essentially derivatives of 4.
Our objective was to develop a synthetic route which
would allow for additional structural variability at C12,
specifically including oxygenation, and to evaluate
alternatives to Wittig/Horner–Emmons methods to
introduce the C13,C14 alkene which could facilitate
* Corresponding author. Tel.: +44 (0)161 200 4530; fax: +001 (510)
217 2371; e-mail: john.gardiner@umist.ac.uk
† For a preliminary report of some of this work, see: Ref. 1.
‡ Visiting researcher from: Departamento de Qu´ımica Farmace´utica,
Facultad de Farmacia, Campus Miguel de Unamuno, Universidad
de Salamanca, 37007 Salamanca, Spain.
Figure 1. Pumiliotoxins and allopumiliotoxins: side-chain
target.
0040-4039/02/$ - see front matter © 2002 Elsevier Science Ltd. All rights reserved.
PII: S0040-4039(02)01071-7

5416
J. M. Gardiner et al. / Tetrahedron Letters 43 (2002) 5415–5418
divergent introduction of groups other than the C14
methyl substituent. These were variations not addressed
in previous syntheses, but of interest given the signifi-
cance of side-chain structure to variations in biological
activity (all natural analogues are C12 methylene, and
syntheses to date do not readily allow variation here).
ring opening would then provide the C11 stereochem-
istry and the terminal hydroxyl present in 5. This was
attractive given the low cost of methacrolein (compared
to the precursor of 6) and the potential to recycle the
unreactive allylic alcohol enantiomer (oxidation, re-
reduction).
Our approach is based on disconnection at the exo-
cyclic alkene yielding 4 since our potential approaches
to the ring system and analogues all involve late forma-
tion of that bond, or its immediate precursor. A C10
aldehyde could also be a precursor of a homologated
acetylene and we reasoned that our syntheses should
proceed through such an intermediate, thus allowing
both for the option to intercept either of the Kibayashi
or Overman routes, as well as being applicable to new
strategies towards the bicyclic ring system and
analogues.11
Thus, crotonaldehyde 11 was subjected to a modified
Corey–Fuchs procedure14 to give the dibromoolefin 12
in 65–70% yield. This was converted to the lithium
acetylide of 8 which was reacted in situ by addition of
methacrolein to give (R,S)-10 directly in ca. 40% yield
over these steps after purification by distillation
(Scheme 2). Sharpless AE kinetic resolution15 led to 9 in
51% yield, 63% e.e. and the slower reacting enantiomer
(R)-10 in 30% yield, 80% e.e. The conditions were not
optimized, but indicated that obtaining 9 with high e.e.
would require considerably lower isolated yields. The
anticipated regiocontrolled epoxide opening with Red-
Al (which has good precedent on comparable sys-
tems),16 which would invert at C11 as required, was
evaluated before pursuing optimization of the kinetic
resolution. The continually low yields for the ring open-
ing led us to abandon this approach, and consequently
we did not further optimize the AE kinetic resolution.
Retrosynthetic analysis of the side-chain backbone
envisages disconnections at C12,C13, with two alterna-
tives for introduction of the C11 stereochemistry and
terminal (C10) alkoxy group (Scheme 1). As an alterna-
tive to addition of 7 we also considered coupling of a
vinyl metal-derived via Negishi zircomethylation of
acetylene 7—to introduce the C13,C14 substituted
alkene directly. One strategy aimed to introduce C11
chirality and C10 hydroxylation from a commercial
chiron starting material; the second approach sought to
evaluate asymmetric resolution of (R,S)-10. Both
approaches ultimately aimed to utilize overall trans
addition of H-Me across keto acetylenes of type 5, via
thermodynamically controlled cuprate addition,12 or
phenylthio conjugate addition and copper catalyzed
Grignard substitution of the phenylthio group.13 Both
routes potentially offered relatively short access to the
fully functionalized side chain with C12 oxygenation,
and to C14 analogues.
We therefore turned to routes utilizing 6 as the C10–
C12 unit, with either attachment of the complete C13–
C17 unit of type 7, already containing the two hydroxy
chiral centres, or of the achiral enyne 8 (subsequently
introducing the C15,C16 functionality). With the syn-
thesis of the acetylide of 8 already established, and the
chiral aldehyde 6 [R1=TBDMS]17 conveniently derived
from methyl (2S)-3-hydroxy-2-methylpropionate (silyla-
tion, then DIBAL-H semi-reduction or by complete
reduction with lithium borohydride and Swern or Dess–
Martin reoxidation), this connectivity was evaluated.
Thus, direct reaction of 6 with the acetylide anion of 8
generated in situ gave the desired enyne 13.18 Yields
after purification were variable (25–50%), with
diastereoselectivity (from crude NMR analysis) of the
order of 4:1. The diastereomeric mixture was oxidized
under Swern conditions to give the ketone 14 in 70%
purified yield. Asymmetric dihydroxylation of 14 using
AD mix-b generated the diol 15, which was then pro
tected as its isopropylidene acetal giving 16 under stan-
The route via (R,S)-10 proposed addition of the
acetylide anion from 8 to methacrolein, followed by
kinetic resolution using Sharpless asymmetric epoxida-
tion to generate 9. We reasoned that directed epoxide
OR
OR
16
15
11
Br
Br
OR
O
R1O
O
OR
ii, iii
R1O
O
i
5
6
7
40%
66%
or
OH
10
11
12
8
4
iv
+
O
OH
OH
8
9
(R)-10
O
OH
OH
9
10
Scheme 2. Reagents and conditions: (i) CBr4, Ph3P, Et3N, 0°C
2.5 h“20°C 1.5 h; (ii) BuLi, HMPA, −78°C; (iii) add
methacrolein“20°C; (iv) Ti(Oi-Pr)4, (+)-DIPT, TBHP,
CH2Cl2, −20°C, 11 h.
Scheme 1. Retrosynthesis of side-chain analogue targets
(R,R1-protecting groups).

J. M. Gardiner et al. / Tetrahedron Letters 43 (2002) 5415–5418
5417
dard conditions (Scheme 3). This provided one route to
the key intermediate type (cf. 5) desired.
moolefin 22.22 Intriguingly, this gave poor reproducibil-
ity in a single-step conversion to acetylene 20, and thus
a two-step route via the E-vinyl bromide was used
(conditions ix). Another approach to 19 via dibro-
moalkene 22 was from the commercially available ethyl
(4S,5R)-2,2,5-trimethyl-1,3-dioxolane-4-carboxylate 24.
Reduction then reoxidation (Dess–Martin periodinane
was also effective) proved more reliable than DIBAL-H
reduction to generate the intermediate aldehyde.
Although the route from commercial 24 is a step
shorter than that from crotonaldehyde (via asymmetric
synthesis from 17), 24 is comparatively expensive and
so particularly on a larger scale, the route from croton-
aldehyde is cost-effective.
Three routes to building blocks of type 7 were evalu-
ated (Scheme 4). The first is analogous to the route in
Scheme 3, but carrying out the asymmetric dihydroxy-
lation prior to introduction of the C12ꢀC13 bond,
through asymmetric dihydroxylation of the TMS
derivative 1719 of enyne 8. Reaction of 17 with AD
mix-b yielded the novel (R,R)-diol 18 in 80% yield and
>90% e.e.20 The diol was protected as its isopropylidene
acetal, and desilylated (both steps >90% yields) to
afford 19.
The same basic unit was also prepared (as cyclohex-
anone-derived acetal 20) by a slightly different route
from PMB-protected crotyl alcohol, elaborated to the
chiral aldehyde 2121 and thence to the novel dibro-
Addition of the lithium acetylide of 19 or 20 to alde-
hyde 6 proceeded to give the adducts 25 and 26, the
best yield (77%) being obtained in the latter case
(Scheme 5). The diastereomers of 26 could be chro-
matographically separated to provide
a
6.7:1
Br
Br
i, ii
iii
diastereomeric ratio. Swern oxidation afforded the
target ketones 16 and 27 in ca. 85% purified yield.
50%
70%
R1O
OH
OH
R1O
O
13
14
There are few methods for the overall trans-addition of
carbon nucleophile-H to either alkynoate or alkynone
conjugate acceptors. Mukaiyama reported overall
trans-addition to alkynoates via a two-step methodol-
ogy employing Z-stereospecific conjugate addition of
sodium thiophenolate to alkynoates, and copper-cata-
lyzed Grignard addition–elimination of the intermedi-
ate vinyl sulfide.13 We had previously evaluated this,
along with thermodynamic cuprate addition to sim-
plified (achiral) analogues of 16 and 27, and ascertained
that use of thiophenol or 2-mercaptothiazole were suc-
cessful in the latter general approach.23 Application of
these conditions to 16 and 27 provided E (28–30) and Z
mixtures in 65–82% total yield, with E:Z ratios around
2:1 and 1:1, respectively. These could be separated
chromatographically most easily for the cyclohexanone-
O
O
iv
v
OH
R1O
O
R1O
O
15
16
Scheme 3. [R1=TBDMS] Reagents and conditions: (i) n-BuLi,
−78°C, 6 h then HMPA; (ii) 6; (iii) (COCl)2, DMSO, Et3N,
−78°C; (iv) AD mix-b, t-BuOH, H2O, 0°C“rt, 20 h; (v) CSA,
(MeO)2CMe2, Me2CO, 12 h.
R
R
HO
O
OH
i
ii, iii
O
TMS
TMS
OHC
R
R
R
R
18
19 R=R=Me
20 R,R=(CH2)5
17
O
O
O
O
i
ii
85%
19 or 20
viii or ix
i, iv-vi
80%
77%
R1O
OH
25 R=R=Me
26 R,R=(CH2)5
R1O
O
O
PMBO
O
16 R=R=Me
27 R,R=(CH2)5
vii
57%
R
R
21
Br
O
O
R
R
R
R
Br
O
O
iii
iv
O
O
O
22 R=R=Me
23 R,R=(CH2)5
x, vi,vii
30-35% (3 steps)
O
24
R1O
O
SAr
R1O
O
EtO2C
28 R=R=Me; Ar=Ph
31 R,R=(CH2)5
Scheme 4. Reagents and conditions: (i) AD mix-b, t-BuOH,
H2O, 0°C“rt, 20 h; (ii) CSA, (MeO)2CMe2, Me2CO, 12 h;
(iii) TBAF; (iv) C6H10O, CSA, THF, rt, 14 h; (v) DDQ,
CH2Cl2, rt, 1 h; (vi) (COCl)2, DMSO, Et3N, −78°C; (vii)
CBr4, Ph3P, Et3N, −20°C, 20 min“20°C, 30 min; (viii) n-
BuLi, 2 equiv., −78°C; (ix) n-BuLi, 1 equiv., −78°C, 10 min,
isolation of vinyl bromide, then KOt-Bu, THF, 2 h; (x)
LiBH4.
29 R,R=(CH2)5; Ar=Ph
30 R,R=(CH2)5; Ar=2-Mer
Scheme
5. [R1=TBDMS;
2-Mer=2-mercaptothiazole]
Reagents and conditions: (i) n-BuLi, −78°C, 2 h then 6“20°C;
(ii) (COCl)2, DMSO, Et3N, −78°C; (iii) ArSH, THF, MeOH,
2 h, rt (for 28/29) or 7 h, D (for 30); (iv) MeMgBr, CuI,
−78°C, 5 h, then −20°C, 13 h“0°C, 1 h.

5418
J. M. Gardiner et al. / Tetrahedron Letters 43 (2002) 5415–5418
derived acetals 29 and 30. Copper catalyzed Grignard
addition–elimination to 29 gave the target 31 in 60%
purified yield.
8. (a) Franklin, A. S.; Overman, L. E. Chem. Rev. 1996, 96,
505–522; (b) Overman, L. E.; Goldstein, S. W.; Rabinow-
itz, M. H. J. Org. Chem. 1992, 57, 1179–1190; (c) Over-
man, L. E.; Robinson, L. A.; Zablocki, J. J. Am. Chem.
Soc. 1992, 114, 368–369.
This thus provides a new route to C12 oxo precursors
to C10–C17 side-chain analogues. C12 is a site of
modification unavailable from natural sources, or prior
chemical syntheses. Reductive ketone removal would
directly afford the natural product side chain, but the
ketone can also be a precursor to reduction to C12-
hydroxy analogues of the side chain. In addition, the
use of a Grignard source for introduction of the
C13,C14 trisubstituted alkene offers the prospect of
future analogue diversification by employing other
alkyl Grignard reagents to access what would constitute
a new class of C14 side-chain analogues.
9. Kibayashi, C.; Aoyagi, S.; Wang, T. C. J. Am. Chem.
Soc. 1993, 115, 11393–11409.
10. (a) Overman, L. E.; Bell, K. L.; Ito, F. J. Am. Chem. Soc.
1984, 106, 4192–4201; (b) Lin, N.-H.; Overman, L. E.;
Rabinowitz, M. H.; Robinson, L. A.; Sharp, M. J.;
Zablocki, J. J. Am. Chem. Soc. 1996, 118, 9062–9072.
11. Gardiner, J. M.; Bruce, S. E. unpublished results.
12. Corriu, R. J. P.; Bolin, G.; Iqbal, J.; Moreau, J. E.;
Vernhet, C. Tetrahedron 1993, 49, 4603–4618 and refer-
ences cited therein.
13. Kobayashi, S.; Mukaiyama, T. Chem. Lett. 1974, 705.
14. Grandjean, D.; Pale, P.; Chuche, J. Tetrahedron Lett.
1994, 35, 3529–3530.
15. Gao, Y.; Hanson, R. M.; Klunder, J. M.; Ko, S. Y.;
Masamune, H.; Sharpless, K. B. J. Am. Chem. Soc. 1987,
109, 5765–5780 (other unsymmetrical 1,1-disubstituted
alkenes gave optimized conditions with up to 95% e.e.).
16. Honda, M.; Katsuki, T.; Yamaguchi, M. Tetrahedron
Lett. 1984, 35, 3857–3860.
Acknowledgements
We thank the British Heart Foundation for Grant
FS/93017 for supporting this work (funding Ph.D. stu-
dentship to P.E.G.), and the Ministerio de Educacio´n,
Cultura y Deporte of Spain for supporting the visit of
M.L.M.M. (2001).
17. (a) Marshall, J. A.; Perkins, J. F.; Wolf, M. A. J. Org.
Chem. 1995, 60, 5556–5559; (b) Paterson, I.; Norcross, R.
D.; Ward, R. A.; Romea, P.; Lister, M. A. J. Am. Chem.
Soc. 1994, 116, 11287–11314.
References
18. As we required oxidation in all systems, separation of
diastereomers was not required, though we did in some
cases chromatographically separate these (see Scheme 5).
For a recent paper in which control in additions to
a-chiral aldehydes, including 6, are discussed and for
further references, see: Evans, D. A.; Allison, B. D.;
Yang, M. G.; Masse, C. E. J. Am. Chem. Soc. 2001, 123,
10840–10852.
19. Farrell, I. W.; Hearn, M. T. W.; Thaller, V. J. Chem.
Soc., Perkin Trans. 1 1978, 1485–1487. We prepared 17
by trapping the Li-acetylide generated from 12 (cf.
Schemes 2 and 3) by addition of TMSCl, and product
purification by Kugelruhr distillation.
1. Gardiner, J. M.; Giles, P. Abstr. Papers Am. Chem. Soc.
1997, 214, 25.
2. (a) Daly, J. W. J. Nat. Prod. 1998, 61, 162–172; (b) Daly,
J. W.; Garraffo, H. M.; Spande, T. F. In Alka-
loids:Chemical and Biological Perspectives; Pelletier, S.
W., Ed.; Elsevier: Oxford, 1999; Vol. 13, pp. 1–161.
3. Daly, J. W.; Myers, C. W. Science 1967, 156, 1970.
4. Daly, J. W.; Tokoyuma, T.; Fujiwara, T.; Highet, R. J.;
Karle, I. L. J. Am. Chem. Soc. 1980, 102, 830.
5. (a) Daly, J. W.; McNeal, E.; Gusovsky, F.; Ito, F.;
Overman, L. E. J. Med. Chem. 1988, 31, 477–480; (b)
Gusovsky, F.; Rossignol, D. P.; McNeal, E.; Daly, J. W.
Proc. Natl. Acad. Sci. USA 1988, 85, 1272–1276; (c) Rao,
K. S.; Warnick, J. E.; Daly, J. W.; Albuquerque, E. X. J.
Pharmacol. Exp. Ther. 1987, 243, 775–783; (d) Daly, J.
W.; Gusovsky, F.; McNeal, E.; Secunda, S.; Bell, M.;
Creveling, C. R.; Nishizawa, Y.; Overman, L. E.; Sharp,
M. J.; Rossignol, D. P. Biochem. Pharm. 1990, 40, 315–
326.
20. Very recently, the (S,S)-diol has been reported. See: Liu,
B.; Chen, M.-J.; Lo, C.-Y.; Liu, R.-S. Tetrahedron Lett.
2001, 42, 2533–2535.
21. e.e. >90% for precursor.
22. The dibromo alkene 22 has been reported only once
previously, on that occasion from L-threonine, but 23 is
unknown. See: Kirschning, A.; Hary, U.; Reis, M. Tetra-
hedron 1995, 51, 2297–2304.
6. Gusovsky, F.; Padgett, W. L.; Creveling, C. R.; Daly, J.
W. Mol. Pharmacol. 1992, 42, 1104–1108.
7. Trost, B. M.; Scanlan, T. S. J. Am. Chem. Soc. 1989, 111,
4988–4990.
23. Gardiner, J. M.; Giles, P. Tetrahedron Lett. 1995, 36,
7519–7522.