Oxazolone cycloadducts as heterocyclic scaffolds for alkaloid construction: Synthesis of (±)-2-epi-pumiliotoxin C

Article abstract of DOI:10.1021/jo902172r

Intramolecular Diels-Alder cycloaddition of N-substituted oxazolone triene I allows direct entry to the functionalized octahydroquinoline II. Further manipulation of this framework by stereo- and regioselective introduction of the 5-methyl substituent, fo

Full text of DOI:10.1021/jo902172r

pubs.acs.org/joc
2, and the related 5β-Z-eneyne variant cis-243A, 3.4 This
Oxazolone Cycloadducts as Heterocyclic Scaffolds
for Alkaloid Construction: Synthesis of
(()-2-epi-Pumiliotoxin C
framework is also found embedded within the tricyclic core
of gephyrotoxin 287C, 4, another related neuroactive alka-
loid.5 Whereas cis-195A has been synthesized on numerous
occasions,6 we were intrigued by the potential offered by the
2-epi-series and sought to develop a general route to compounds
of this ilk. In this paper we describe one such route, culminating
in a preliminary synthesis of 2-epi-pumiliotoxin C, 5.
Stephen Philip Fearnley* and Charnsak Thongsornkleeb†
Department of Chemistry, The City University of New York -
The Graduate Center & York College, 94-20 Guy R. Brewer
Boulevard, Jamaica, New York, 11451. †Current address:
Laboratory of Organic Synthesis, Chulabhorn Research
Institute, 54 Moo 4 Vibhavadee-Rangsit Highway, Lak Si,
Bangkok 10210, Thailand.
fearnley@york.cuny.edu
Received October 8, 2009
FIGURE 1. Representative decahydroquinoline targets.
As part of an ongoing program of research at the interface
of heterocyclic and natural products chemistry, we recently
reported the first intramolecular Diels-Alder reaction of
N-alkylated oxazolones.7 This thermally activated cycloaddi-
tion allows rapid entry into a number of important hetero-
cyclic frameworks, chiefly, densely functionalized hexahy-
droindoles and octahydroquinolines. As we continued our
studies of oxazolone chemistry,8 it was clear that the latter
appeared ideally suited to construction of decahydroquino-
line alkaloids via direct incorporation of the initial cycload-
duct framework. We therefore selected 2-epi-pumiliotoxin C,
5, by way of example. Although previously synthesized on a
number of occasions,9 most often as a minor byproduct en
route to pumiliotoxin C itself, this compound has remained
an important benchmark for the testing of new synthetic
methodologies and is thus a fitting initial target.
Intramolecular Diels-Alder cycloaddition of N-substi-
tuted oxazolone triene I allows direct entry to the func-
tionalized octahydroquinoline II. Further manipulation
of this framework by stereo- and regioselective introduc-
tion of the 5-methyl substituent, followed by excision of
the carbamate, yields (()-2-epi-pumiliotoxin C.
The skin secretions of neotropical frogs have long proven a
rich and varied source of alkaloid natural products, many of
which possess powerful yet specific bioactivity allied to their
complex structure. A case in point is provided by the decahy-
droquinolines, with over 50 examples reported to date.1
Several of these mild toxins display intriguing neurological
activity as reversible antagonists of the nicotinic acetylcholine
receptor channel.2 As many of the proposed structures are
based solely on MS and IR studies, confirmation by synthesis
provides further compelling reason for development of a gene-
ral approach to targets of this type. The parent member of this
class, cis-195A3 (the alkaloid formerly known as pumiliotoxin
C, 1), is typical of their general 2,5-disubstituted cis-fused
decahydroquinoline pattern, Figure 1. Altogether rarer is the
2-epi-cis motif, represented by 2β,5β-diallyl congener cis-219A,
In keeping with the overall theme of our methodology,
our proposed synthesis is based on a key intramolecular
(4) Tokuyama, T.; Tsujita, T.; Shimada, A.; Garraffo, H. M.; Spande, T.
F.; Daly, J. W. Tetrahedron 1991, 47, 5401–5414.
(5) Daly, J. W.; Witkop, B.; Tokuyama, T.; Nishikawa, T.; Karle, I. L.
Helv. Chim. Acta 1977, 60, 1128–1140.
€
(6) (a) For the first reported synthesis, see: Oppolzer, W.; Frostl, W.;
Weber, H. P. Helv. Chim. Acta 1975, 58, 593–595. (b) For a recent synthesis,
see: Girard, N.; Hurvois, J.-P.; Moinet, C.; Toupet, L. Eur. J. Org. Chem.
2005, 2269–2280.
(7) Fearnley, S. P.; Market, E. Chem. Commun. 2002, 438–439.
(8) For reviews, see: (a) Fearnley, S. P. Current Organic Chemistry; Vol. 8,
Asymmetric Synthesis; Atta-Ur-Rahman, Ed.; Bentham Science Publishers: The
Netherlands, 2004; pp 1289-1338; (b) Matsunaga, H.; Ishizuka, T.; Kunieda, T.
Tetrahedron 2005, 61, 8073–8094.
(1) Daly, J. W.; Spande, T. F.; Garraffo, H. M. J. Nat. Prod. 2005, 68,
1556–1575.
(2) Warnick, J. E.; Jessup, P. J.; Overman, L. E.; Eldefrawi, M. E.; Nimit,
Y.; Daly, J. W.; Albuquerque, E. X. Mol. Pharmacol. 1982, 22, 565–573.
Daly, J. W.; Nishizawa, Y.; Padgett, W. L.; Tokuyama, T.; McCloskey, P. J.;
Waykole, L.; Schultz, A. G.; Aronstam, R. S. Neurochem. Res. 1991, 16,
1207–1212.
(9) (a) Oppolzer, W.; Flaskamp, E.; Bieber, L. W. Helv. Chim. Acta 2001,
84, 141–145. (b) Riechers, T.; Krebs, H. C.; Wartchow, R.; Habermehl, G.
Eur. J. Org. Chem. 1998, 2641–2646. (c) Back, T. G.; Nakajima, K. J. Org.
Chem. 1998, 63, 6566–6567. (d) Back, T. G.; Nakajima, K. Tetrahedron Lett.
1997, 38, 989–992. (e) Meyers, A. I.; Milot, G. J. Am. Chem. Soc. 1993, 115,
6652–6660. (f) Polniaszek, R. P.; Dillard, L. W. J. Org. Chem. 1992, 57, 4103–
4110.
(3) Daly, J. W.; Tokuyama, T.; Habermehl, G.; Karle, I. L.; B. Witkop, B.
Liebigs Ann. Chem. 1969, 727, 198–204.
DOI: 10.1021/jo902172r
r
Published on Web 12/29/2009
J. Org. Chem. 2010, 75, 933–936 933
2009 American Chemical Society

Fearnley and Thongsornkleeb
SCHEME 3. Initial Cycloaddition Studies
JOCNote
SCHEME 1. Oxazolone IMDA Approach to 2-epi-Pumilio-
toxin C
TABLE 1. Reaction Conditions vs Selectivity
conditions (in o-dichlorobenzene)
yield (%)
6:14
sealed tube, 210 °C, 72 h
microwave, 225 °C, 5 h
1.3 equiv of EtAlCl2, reflux, 72 h
78
59
57
3.1:1
2.5:1
1:0
three-step protocol as developed by Shibuya12 was utilized to
introduce the secondary oxazolone moiety: Mitsunobu reac-
tion with oxazolidine-2,4-dione 12 followed by reduction
and elimination to generate the dienophilic olefin.
SCHEME 2. Synthesis of Triene Precursor
Investigation of the key cycloaddition ensued (Scheme 3).
Simple thermal activation proceeded smoothly to furnish two
main cycloadducts, 6 and 14, in 3.7:1 ratio and 81% overall
yield. Ring junction stereochemistry was readily assigned
as cis and trans for each, in accordance with our previous
findings. However, the signal at C2 was initially occluded, and
extensive decoupling studies were required.13 These indeed
confirmed the C2-propyl substituent as axial in both cycload-
ducts, with added proof for 6 from X-ray crystallography.
Considerableeffort was spentonoptimizationof conditions; a
brief survey is presented in Table 1 for comparison. In all
cases, C2-epi-cis-cycloadduct 6 predominated, even to the
extent that under thermal Lewis acid-promoted conditions14
it was the sole product isolated.
With the structure of the major cycloadduct confirmed as
6, we proceeded with the next phase of our approach. This
centered on successful installation of the 5-methyl substitu-
ent in regio- and stereocontrolled fashion. In order to
address both these concerns simultaneously, we chose to
pursue a two-step protocol - cyclopropanation followed by
reductive cleavage. Given the inherent topology of the cyclo-
adduct, we reasoned that simple cyclopropanation would
favor the convex β-face. Subsequent hydrogenolysis of the
most sterically accessible edge would locate the 5-methyl as
required. If successful, this would offer a fairly novel solution
for stereocontrolled installation of an isolated methyl sub-
stituent, though precedent does arise from a seminal study on
bicyclo[4.1.0]heptane.15
In the event, cyclopropanation under modified Denmark
conditions16 proceeded with a high degree of β-facial control
to yield the expected tetracycle 1513 (Scheme 4). Subsequent
rupture at the 6-position was best achieved by simple hydro-
genation over Adams’ catalyst,17 with reasonable selectivity
in favor of the desired isomer 16 (5.5:1, 5-Me to 6-Me).
However, for reasons yet unclear to us this transformation
required superstoichiometric quantities of PtO2 to effect
Diels-Alder cycloaddition, the reaction of an oxazolone die-
nophile with a substituted diene fragment, tethered via nitro-
gen. This first-generation approach involves trienic system 8, as
shown in our initial retrosynthetic analysis (Scheme 1). Our
earlier studies with the heptadienyl parent system had shown
the cis-cycloadduct to predominate [3.8:1, cis to trans] via
an endo mode of cycloaddition.7 Assuming a similar transition
state, the precise aspect of the propyl substituent would now be
the critical issue. From early examination of models, we
surmised that a pseudoaxial arrangement would minimize
unfavorable allylic-type strain10 with the oxazolone carbonyl.
Cycloaddition in this fashion would thus establish three of the
four required stereocenters in a single step. It would only
remain to introduce the 5-methyl group in regio- and stereo-
selective fashion and then excise the carbamate functionality,
most probably by a hydrolysis-deoxygenation protocol.
Construction of the required triene system proceeded
from commercially available divinyl carbinol 9 (Scheme 2).
Saucy-type rearrangement11 afforded heptadienal 10, which
underwent simple Grignard reaction to yield alcohol 11. A
(13) See the Supporting Information for details of structural assignment.
(14) Martin, S. F.; Rein, T.; Liao, Y. Tetrahedron Lett. 1991, 32, 6481–
6484.
(15) Stahl, K.-J.; Hertzsch, W; Musso, H. Liebigs Ann. Chem. 1985, 1474–
1484.
(10) Hoffmann, R. W. Chem. Rev. 1989, 89, 1841–1860. For specific
examples of R-substituents exerting similar stereocontrol in IMDA of N-acyl
€
systems, see: Friedrich, A.; Jainta, M.; Nising, C. F.; Brase, S. Synlett 2008,
589–591. and ref 6a.
(11) Reed, S. F. J. Org, Chem. 1965, 30, 1663–1665.
(12) Yuasa, Y.; Ando, J.; Shibuya, S. J. Chem. Soc., Perkin Trans. 1 1996,
465–473.
(16) Denmark, S. E.; Edwards, J. P. J. Org. Chem. 1991, 56, 6974–6981.
(17) Rigby, J. H.; Laxmisha, M. S.; Hudson, A. R.; Heap, C. H.; Heeg, M.
J. J. Org. Chem. 2004, 69, 6751–6760.
934 J. Org. Chem. Vol. 75, No. 3, 2010

Fearnley and Thongsornkleeb
JOCNote
SCHEME 4. Synthesis of (()-2-epi-Pumiliotoxin C
Experimental Section
Diels-Alder Cycloaddition of Triene 8. To a stirred solution of
triene 8 (698 mg; 3.16 mmol; 1.0 equiv) in o-dichlorobenzene
(63.1 mL; ∼0.05 M) was added a trace of hydroquinone (14 mg;
2% by wt). The colorless solution was freeze-thaw degassed
(3ꢀ), placed under an atmosphere of argon, and brought rapidly
to reflux. After 72 h, heat was removed and the brown solution
partially reduced in vacuo (hi-vac; 85 °C) to yield 878 mg of the
crude cycloadducts as a dark brown solid. Further purification
by flash chromatography (silica ratio 100:1; hexane/EtOAc, 4:1
to 3:1 to 2:1 to 1:1) yielded each cycloadduct as a colorless solid.
First, the minor trans-cycloadduct, (()-(31S,4R,6aR,9aR)-4-
propyl-4,5,6,6a,9,9a-hexahydrooxazolo[5,4,3-ij]quinolin-2(31H)-
one, 14: 120 mg; 0.54 mmol; 17%; Rf 0.36 (hexane/EtOAc 2:1);
1H NMR (400 MHz; CDCl3) δ 5.77 (m, 2H), 4.58 (td, J = 2 ꢀ
8.1, 5.9 Hz, 1H), 3.96 (dt, J = 10.1, 2 ꢀ 5.1 Hz, 1H), 2.99 (dd,
J = 10.0, 8.0 Hz, 1H), 2.89 (m, 1H), 2.20 (ddm, J = 14.2, 5.8 Hz,
1H), 1.94 (dq, J = 12.9, 3 ꢀ 3.3 Hz, 1H), 1.81 (m, 1H), 1.75 (ddd,
J = 13.4, 5.8, 4.2 Hz, 1H), 1.61-1.71 (m, 2H), 1.58 (qd, J = 3 ꢀ
12.8, 3.2 Hz, 1H), 1.29-1.43 (m, 3H), 0.92 (t, J = 7.2 Hz, 3H);
13C NMR (DEPT) (100.6 MHz; CDCl3) δ 157.6 (C), 132.0
(CH), 125.4 (CH), 71.6 (CH), 57.2 (CH), 49.8 (CH), 37.0 (CH),
33.2 (CH2), 28.7 (CH2), 28.4 (CH2), 23.4 (CH2), 19.5 (CH2), 13.7
(CH3); IR (thin film) νmax 2958, 2932, 1744, 1020 cm-1; HRMS
m/z [M þ H]þ calcd for C13H20NO2 222.1494, found 222.1489.
Next, the major cis-cycloadduct, (()-(31S,4R,6aS,9aR)-4-pro-
pyl-4,5,6,6a,9,9a-hexahydrooxazolo[5,4,3-ij]quinolin-2(31H)-one,
6: 444.4 mg; 2.01 mmol; 64%; Rf 0.22 (hexane/EtOAc 2:1); 1H
NMR (400 MHz; CDCl3) δ 5.91 (ddt, J = 9.5, 6.4, 2 ꢀ 3.1 Hz,
1H), 5.67 (dm, J = 9.5 Hz, 1H), 4.88 (dddd, J = 9.3, 4.0, 1.9, 1.2
Hz, 1H), 4.04 (ddd, J = 9.3, 4.6, 1.4 Hz, 1H), 3.88 (dt, J = 9.6,
2 ꢀ 4.8 Hz, 1H), 2.59 (ddd, J = 16.3, 7.0, 2.0 Hz, 1H), 2.13 (m,
1H), 2.02 (ddq, J = 16.4, 3.9 Hz, 3 ꢀ 2.9 Hz, 1H), 1.90 (tt, J =
2 ꢀ 13.9, 2 ꢀ 3.8 Hz, 1H), 1.64-1.80 (m, 3H), 1.43 (dm, J = 13.9
Hz, 1H), 1.26-1.39 (m, 3H), 0.92 (t, J = 7.2 Hz, 3H); 13C NMR
(DEPT) (100.6 MHz; CDCl3) δ 157.2 (C), 131.4 (CH), 126.7
(CH), 72.4 (CH), 51.9 (CH), 48.5 (CH), 31.8 (CH2), 30.9 (CH),
27.6 (CH2), 23.3 (CH2), 22.6 (CH2), 19.4 (CH2), 13.8 (CH3); IR:
(thin film) νmax 3024, 2943, 2867, 1731, 1047 cm-1; HRMS m/z
[M þ H]þ calcd for C13H20NO2 222.1494, found 222.1487. Anal.
Calcd for C13H19NO2: C, 70.56; H, 8.65; N, 6.33. Found: C,
70.51; H, 8.64; N, 6.24.
completion. The intermediacy of a discrete platinocyclic
species has not been discounted.18
As this initial 5.5:1 mix proved inseparable, it was em-
ployed directly in studies of oxazolidinone cleavage. We
found methyl lithium19 to be doubly effective in this regard,
as not only did this afford the amine conveniently protected
as acetamide, but also enabled separation of the required
regioisomer 17, in 79% overall yield from 15. Once again
X-ray crystallography provided unequivocal confirmation
of our ongoing structural assignments. Deoxygenation at C8
proceeded smoothly via a modified Barton-McCombie
protocol20 to afford the N-acetylated target, 19.
In these latter stages of the synthesis (17-19), both 1H and
13C spectra were complicated by the presence of broadened
and/or doubled signals. This observation was also noted by
Meyers,9e and studied in detail by Polniaszek9f and Daly,21
all of whom concluded this arose from rapid equilibration of
the cis-fused bicyclic systems, in itself a well-established
property of cis-decahydroquinolines.22,23 A final deprotec-
tion of the N-acetamide group under Pearson’s dissolving
metal conditions24 yielded (()-2-epi-pumiliotoxin C, 5, iden-
tical in all respects to that previously reported.9a,c,e
Direct Cyclopropanation of Major Cycloadduct: (()-(3R,41S,
6aR,7aR,8aR,8bS)-3-Propyloctahydro-1H-cyclopropa[f]oxazolo-
[5,4,3-ij]quinolin-5(41H)-one, 15. A 75-mL heavy-walled sealed
tube, fitted with septum and stir-bar, was flame-dried under
vacuum, flushed with Ar, and allowed to cool to rt. CH2Cl2
(2 mL) was added, followed by diethylzinc (Aldrich, 1.0 M in
heptane; 2.65 mL; 2.65 mmol; 5.0 equiv) and the solution cooled
to 0 °C. Neat chloroiodomethane (0.39 mL; 5.36 mmol; 10.1
equiv) was added dropwise with caution over a period of 2 min in
order to prevent exotherm. During addition, a heavy colorless
gum was precipitated; this was broken up by addition of further
CH2Cl2 (5 mL), followed by vigorous stirring, to yield a milky
suspension. The mixture was allowed to warm to rt. After
10 min, a solution of cis-cycloadduct 6 (117 mg; 0.53 mmol;
1.0 equiv) in CH2Cl2 (2 mL þ 2 mL rinse) was added dropwise.
The tube was sealed and heated immediately to 120 °C with
vigorous stirring. After 5 h, heat was removed and the reaction
allowed to cool to rt. The tube was opened, the contents poured
onto 1 M HCl (25 mL), and the residue broken up with further 1
M HCl (25 mL) and CH2Cl2 (25 mL). The resulting mixture was
partitioned, and the aqueous phase extracted with further
CH2Cl2 (2 ꢀ 25 mL). The combined organic phases were dried
(MgSO4), filtered, and reduced in vacuo to yield 161 mg of crude
cyclopropanes as a yellow oil. Further purification by flash
chromatography (silica ratio 300:1; hexane/EtOAc 4:1 to 3:1)
In summary, this study has demonstrated the utility
of oxazolone IMDA cycloadducts as valuable scaffolds for
alkaloid construction, culminating in the synthesis of (()-2-
epi-pumiliotoxin C by a direct incorporation approach.
(18) Jennings, P. W.; Johnson, L. L. Chem. Rev. 1994, 94, 2241–2290.
(19) Tius, M. A.; Kerr, M. A. J. Am. Chem. Soc. 1992, 114, 5959–5966.
(20) (a) Imidazoylthiocarbonyl formation: Winbush, S-A. M.; Mergott,
D. J.; Roush, W. R. J. Org. Chem. 2008, 73, 1818–1829. (b) Deoxygenation:
Chochrek, P.; Wicha, J. J. Org. Chem. 2007, 72, 5276–5284.
(21) Spande, T. F.; Jain, P.; Garraffo, H. M.; Pannell, L. K.; Yeh, H. J. C.;
Daly, J. W. J. Nat. Prod. 1999, 62, 5–21.
(22) Booth, H.; Bostock, A. H. J. Chem. Soc., Perkin Trans. 2 1972, 615–
621.
(23) Other possible complications include rotamers of the acetamide and
quadrupolar broadening by nitrogen. We are grateful to the two reviewers
who suggested these additional phenomena.
(24) Pearson, A. J.; Rees, D. C. J. Chem. Soc., Perkin Trans. 1 1982, 2467–
2476.
J. Org. Chem. Vol. 75, No. 3, 2010 935

Fearnley and Thongsornkleeb
JOCNote
yielded the title β-cyclopropane 15 as a colorless oil: 107 mg; 0.46
mmol; 86%; Rf 0.25 (hexane/EtOAc 2:1); 1H NMR (400 MHz;
CDCl3) δ 4.72 (dt, J = 9.2, 2 ꢀ 2.6 Hz, 1H), 3.98 (dt, J = 10.3,
2 ꢀ 5.2 Hz, 1H), 3.72 (dd, J = 9.4, 3.8 Hz, 1H), 2.51 (ddd, J =
15.3, 6.7, 2.9 Hz, 1H), 1.95 (m, 1H), 1.86 (m, 2H), 1.71 (m, 1H),
1.43 (dm, J = 12.8 Hz, 1H), 1.29-1.40 (m, 3H), 1.09 (app sextet,
J = 3.6 Hz, 1H), 0.94 (t, J = 7.0 Hz, 3H), 0.91-0.97 (occluded
m, 1H), 0.87 (td, J = 2 ꢀ 8.2, 4.4 Hz, 1H), 0.82 (ddd, J = 15.5,
8.4, 2.3 Hz, 1H), 0.73 (qd, J = 3 ꢀ 8.2, 4.3 Hz, 1H), -0.03 (q, 4.3
Hz, 1H); 13C NMR (DEPT) (100.6 MHz; CDCl3) δ 156.6 (C),
73.1 (CH), 51.5 (CH), 48.4 (CH), 32.4 (CH), 32.2 (CH2), 27.5
(CH2), 22.6 (CH2), 22.1 (CH2), 19.5 (CH2), 15.4 (CH2), 13.8
(CH3), 6.7 (CH), 4.4 (CH); IR (thin film) νmax 2933, 2868, 1739,
1051 cm-1; HRMS m/z [M þ H]þ calcd for C14H22NO2
236.1651, found 236.1641.
Regioselective Hydrogenolysis of Cyclopropane Ring:
(()-(31S,4R,6aS,7R,9aR)-7-Methyl-4-propyloctahydrooxazolo-
[5,4,3-ij]quinolin-2(31H)-one, 16. To a stirred solution of β-
cyclopropane 15 (123 mg; 0.52 mmol; 1.0 equiv) in EtOAc (15
mL), cooled to 0 °C, was added platinum oxide (95 mg; 0.42
mmol; 0.8 equiv) and the dark brown suspension evacuated--
flushed with H2 (3ꢀ). Cooling was removed and stirring main-
tained under an atmosphere of H2 (1 atm) at rt. After 8 h,
additional PtO2 (48 mg; 0.21 mmol; 0.4 equiv) was added and
stirring maintained under H2. Further PtO2 (2 ꢀ 0.4 equiv) was
added at 24 h, and again at 33 h, for a total of 2.0 equiv. At each
addition, the suspension would darken from brown to black and
then precipitate a fine metallic solid, which could be resus-
pended upon vigorous stirring. After 48 h total, the suspension
was filtered through Celite (1 cm ꢀ 3 cm plug), flushing with
EtOAc (3 ꢀ 10 mL). The combined organic phases were reduced
in vacuo to yield an inseparable mixture of the desired 5-methyl
isomer 16, along with the minor 6-isomer (5.5 to 1 by NMR):
120 mg. This was used directly in the ensuing step without fur-
ther purification: Rf 0.37 (hexane/EtOAc 2:1); 1H NMR (400 MHz;
CDCl3) (major isomer) δ 4.52 (dt, J = 7.6, 2 ꢀ 7.2 Hz, 1H), 3.84
(qm, J = 7.2 Hz, 1H), 3.80 (dd, J = 7.2, 4.4 Hz, 1H), 1.98 (m,
1H), 1.78 (dq, J = 13.6, 3 ꢀ 2.9 Hz, 1H), 1.25-1.72 (m, 12H
total), 0.92 (t, J = 7.4 Hz, 3H), 0.91 (d, J = 6.4 Hz, 3H).
Final Deprotection to (()-2-epi-Pumiliotoxin C: (2R,4aS,5R,
8aR)-5-Methyl-2-propyldecahydroquinoline, 5. A two-necked
25 mL flask, equipped with septum, stir bar, dry ice coldfinger
condenser, and bubbler, was flushed with liquid NH3 (∼9 mL)
and allowed to reflux (-33 °C). To this was added, dropwise, a
cosolution of acetamide 19 (26.7 mg; 0.112 mmol; 1.0 equiv) and
anhydrous EtOH (18.5 mL; 0.317 mmol; 2.82 equiv) in DME (1.03
mL), followed by calcium (32.5 mg; 0.812 mmol; 7.22 equiv),
resulting in a dark blue solution. After 3.5 h under reflux, excess
calcium was quenched by addition of EtOH (2 mL) and the
ammonia allowed to evaporate. The resulting white slurry was
taken up in CHCl3 (10 mL) and water (10 mL), and the pH of the
aqueous layer adjusted from pH 14 to 1 using 6 M HCl. Sat. aq.
K2CO3 (10 mL) was added, resulting in a flocculent colorless
precipitate. The biphasic mixture was filtered, rinsing with addi-
tional CHCl3 (5 mL), partitioned, and the aqueous phase extracted
with further CHCl3 (2 ꢀ 10 mL). The combined organic phases
were dried (K2CO3), filtered, and reduced in vacuo to yield 2-epi-
pumiliotoxin C 5 as a colorless oil: 14.3 mg; 0.073 mmol; 65%; Rf
0.05 (hexane/EtOAc 1:2); 1H NMR (400 MHz; CDCl3) δ 3.06 (dt,
J = 10.5, 2 ꢀ 4.3 Hz, 1H), 2.75 (dm, J = 6.0 Hz, 1H), 1.61-1.84
(m, 4H total), 1.21-1.56 (m, 11H total), 0.99-1.13 (m, 2H total),
0.95 (d, J = 7.2 Hz, 3H), 0.86 (t, J = 6.8 Hz, 3H); 13C NMR (100.6
MHz; CDCl3) δ 50.0, 49.7, 42.2, 38.4 (br), 32.5 (br), 31.4 (br), 28.8
(v. br - 2 signals), 25.3 (br), 20.6, 19.4, 19.3, 14.1; IR (thin film)
ν
max 2926, 2868, 1461, 1376 cm-1; HRMS m/z [M þ H]þ calcd for
C13H26N 196.2065, found 196.2058.
Acknowledgment. We are grateful to the National Insti-
tutes of Health (GM08153-30) for generous support of this
research. Acknowledgment is also made to the Donors of the
American Chemical Society Petroleum Research Fund for
partial support. We thank Dr. Cliff Sol of the CUNY Mass
Spectrometry Facility, and Dr. Louis Todaro of the CUNY
X-ray Facility, both at Hunter College, for mass spectral and
X-ray crystallographic analysis, respectively. These facilities
are supported by a “Research Centres in Minority Institu-
tions” award, RR-03037, from NCRR/NIH. Funding for
MS instrumentation was provided by an NIH Shared In-
strumentation Grant (1S10RR022649-01), and the CUNY
Instrumentation Fund.
Supporting Information Available: Full experimental pro-
C
cedures and data for all remaining new compounds; 1H and 13
NMR spectra for compounds 8, 6, 14, 15-19, and 5; solved X-
ray structures for 6 and 17. This material is available free of
charge via the Internet at http://pubs.acs.org.
936 J. Org. Chem. Vol. 75, No. 3, 2010