A concise synthesis of (±) and a total synthesis of (+)-epiquinamide

Article abstract of DOI:10.1016/j.tetlet.2006.05.092

A total synthesis of the quinolizidine alkaloid (+)-epiquinamide 1 has been achieved starting from (-)-pipecolinic acid 3. The key step is the highly diastereoselective addition of a TBDMS-protected propargyl alcohol to a chiral aldehyde derived from 3 to give erythro alkynol 19, which is then easily transformed into the desired bicyclic skeleton.

Full text of DOI:10.1016/j.tetlet.2006.05.092

Tetrahedron Letters 47 (2006) 5017–5020
A concise synthesis of ( ) and a total synthesis of (+)-epiquinamide
Sok Teng (Amy) Tong and David Barker*
Department of Chemistry, University of Auckland, 23 Symonds St., Auckland, New Zealand
Received 5 April 2006; revised 10 May 2006; accepted 17 May 2006
Available online 6 June 2006
Abstract—A total synthesis of the quinolizidine alkaloid (+)-epiquinamide 1 has been achieved starting from (ꢀ)-pipecolinic acid 3.
The key step is the highly diastereoselective addition of a TBDMS-protected propargyl alcohol to a chiral aldehyde derived from 3
to give erythro alkynol 19, which is then easily transformed into the desired bicyclic skeleton.
ꢀ 2006 Elsevier Ltd. All rights reserved.
Epiquinamide 1 is a quinolizidine alkaloid recently iso-
lated from extracts of the skin of Epipedobates tricolor,
an Ecuadorian poisonous frog.1 Epiquinamide 1, has
been found to be highly selective for b2 nicotinic acetyl-
choline receptors (nAChRs), as such, representing a new
structural class of nicotinic agonists and could be
considered a lead compound for the development of
nAChR therapeutic agents. The minute amount
(ꢁ240 lg) of 1 isolated from the skin extracts was
enough to determine the structure and the relative ste-
reochemistry of the natural product as (1R*,10R*)-1-ace-
tamidoquinolizidine. Due to 1 being a novel nicotinic
agonist with unresolved absolute stereochemistry it
was decided to synthesise epiquinamide 1 with the aim
of developing new nicotinic agents. During the course
of this work, the structure and relative stereochemistry
of epiquinamide 1 were confirmed by two independent
asymmetric syntheses.2,3 We herein report a simple syn-
thesis of ( )-epiquinamide 1 and its (1S*,10R*) diaste-
reoisomer 2 as well as the synthesis of (+)-1, both
syntheses utilising commercially available pipecolinic
acid 3 as the starting material.
pipecolinic acid 3 was converted into ethyl pipecolinate
hydrochloride8 using thionyl chloride in ethanol and
then alkylated with ethyl 4-bromobutyrate to give dies-
ter 69,10 in 85% overall yield. Dieckmann cyclisation of
diester 6 was achieved using potassium tert-butoxide in
THF at room temperature to give keto-ester 77,10 in high
yield, the NMR spectra of which showed that it existed
exclusively in the enol form. Hydrolysis and decarboxyl-
ation of 7 was achieved by heating in 4 M HCl over-
night11 to give ketone 86 in 89% yield; it was found
that this method was higher yielding and more reliable
than refluxing in sulfuric acid.6,9 Ketone 8 was found
to be particularly unstable (ꢁ50% decomposition oc-
curred if left standing overnight at room temperature)
and was used immediately. Conversion of ketone 8 to
oxime 5 was achieved in 82% yield, using the previously
reported method.6 Reduction of oxime 5 using LiAlH4
in THF gave
a
1.1:1 ratio of (1S*,10S*)-4 to
(1R*,10S*)-4, in 84% yield, which were separated using
repeated column chromatography (98:2 MeOH/aq
NH3). Both amines 4 were acetylated using acetyl chlo-
ride in DMF to give ( )-epiquinamides 1 and 2 in rea-
sonable yields. The 1H NMR spectrum of 1 was
identical to that reported for the isolated natural
product.1
For confirming the relative stereochemistry of 1 we
decided to prepare both diastereoisomers of 1-amino-
quinolizidine 4, which have been previously reported
only as a mixture of diastereoisomers,4–6 from the reduc-
tion of the known oxime 5.6,7 During our synthesis of
oxime 5, we found many of the previously reported steps
to be low yielding and irreproducible, and hence it was
decided to optimise this synthesis (Scheme 1). Racemic
Disappointed with the ratio of products from the oxime
reduction, we investigated an alternative method of
introducing the 1-amino group. Reduction of ketone 8
with NaBH4 gave a 5:1 ratio of the desired equatorial
alcohol 912–14 over the axial 10, and that the separation
of the two alcohols by column chromatography (using
100% MeOH) being far more easily achieved than sepa-
ration of amines 4. Alcohol 9 was then converted via the
mesylate to azide 11 in 75% overall yield. The optimal
*
Corresponding author. Fax: +64 9 373 7422; e-mail: d.barker@
auckland.ac.nz
0040-4039/$ - see front matter ꢀ 2006 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tetlet.2006.05.092

5018
S. T. Tong, D. Barker / Tetrahedron Letters 47 (2006) 5017–5020
OH
O
CO2Et
CO2Et
CO2H
(iii)
(iv)
(i), (ii)
N
CO2Et
NHR
N
NH
N
8
3
6
7
(v)
OH
N
NHR
H
H
N
(vi)
+
N
N
5
R=H 4b
R=Ac ( )-2
R=H 4a
R=Ac ( )-1
(vii)
(vii)
Scheme 1. Reagents, conditions and yields: (i) SOCl2, EtOH, reflux, 2 h, 99%; (ii) 1.05 equiv ethyl 4-bromobutyrate, K2CO3, acetone, reflux, 22 h,
85%; (iii) 2 equiv KOtBu, THF, 0 ꢁC to rt, 2 h, 95%; (iv) 4 M HCl, 90–100 ꢁC, 12 h, 89%; (v) NH2OHÆHCl, pyridine, EtOH, reflux, 1 h, 82%; (vi)
LiAlH4, THF, 0 ꢁC to rt, 1 h, 4a (45%) and 4b (39%); (vii) AcCl, NEt3, DMF, rt, 20 h, 1 (69%), 2 (72%).
temperature for azide addition was found to be 70–
80 ꢁC, below which only trace amounts azide 11 was
formed and unreacted mesylate was returned whilst tem-
peratures above 100 ꢁC resulted in extensive decomposi-
tion. Reduction of azide 11 was achieved in 92% yield
using LiAlH4 to give amine 4, which was acetylated as
above to give ( )-1 (Scheme 2).
CO2H
CO2Me
(i), (ii)
NH
N
3
12
(iii), (iv)
O
OH
(v)
With racemic 1 in hand we then wished to synthesize 1
asymmetrically, again beginning with pipecolinic acid
3. Starting from chirally pure 3 using the schemes
outlined above was, however, unfeasible as complete
racemisation occurred during the Dieckmann and
decarboxylation steps of the synthesis. We therefore
turned to alternative methods to convert pipecolinic acid
3 to the required quinolizidine ring system and we ini-
tially envisaged using ring-closing metathesis to form
the second ring and began by trialing the route on race-
mic 3 (Scheme 3). Pipecolinic acid 3 was converted to the
methyl ester and then allylated to form 1215 in 65% over-
all yield. Reduction of ester 12 with LiAlH4 gave the pri-
mary alcohol which was then oxidized using Swern
conditions16,17 to give the unstable aldehyde 1318 in a
modest 50% yield. It was discovered that the use of a
workup procedure with ammonium chloride17 resulted
N
N
15
13
(vi)
OH
N
S
N
N
14
16
Scheme 3. Reagents, conditions and yields: (i) SOCl2, MeOH, reflux,
2 h, 97%; (ii) 1 equiv allyl bromide, K2CO3, acetone, reflux, 17 h, 68%;
(iii) LiAlH4, THF, 0 ꢁC to rt, 50 min, 92%; (iv) (COCl)2, DMSO,
DCM, ꢀ78 ꢁC to rt, 1 h, 50%; (v) 4 equiv vinyl magnesium bromide,
0 ꢁC to rt, 2 h, 66% (1:1 mixture of inseparable diastereoisomers); (vi)
Grubb’s first and second generation RCM catalysts (5 and 10 mol %)
DCM, 16 (0%).
OH
in significant amounts of methylthiomethyl imine 14
being generated, which was alleviated by the use of brine
rather than the ammonium chloride. Addition of vinyl
magnesium bromide to aldehyde 13 gave alcohol 15 as
a 1:1 inseparable mixture of diastereoisomers in 66%
yield. Unfortunately ring-closing metathesis of diene
15, using either Grubb’s first or second generation cata-
lysts under a variety of conditions, did not result in the
desired allylic alcohol 16. Decomposition of the starting
material occurred under all the conditions tested and
this route was abandoned.
O
OH
H
H
(i)
+
N
N
N
8
10
9
(ii)
NHAc
N3
H
N
H
(iii, iv)
N
11
( )-1
We then turned to an approach involving addition of a
nucleophile with all the required carbons to form the
second ring to the known aldehyde 17.19,20 Aldehyde
17 was synthesized from S-3, which in turn was obtained
by resolution of the tartrate salts of racemic 3.21 Addi-
Scheme 2. Reagents, conditions and yields: (i) NaBH4, MeOH, 0 ꢁC to
rt, 6 h, 10 (16%) and 9 (82%); (ii) (a) 1.2 equiv MsCl, NEt3, DCM,
0 ꢁC, 30 min; (b) 5 equiv NaN3, DMF, 70–80 ꢁC, 18 h, 75% from 9; (iii)
LiAlH4, THF, rt, 3 h, 92%; (iv) AcCl, NEt3, DMF, rt, 21 h, 73%.

S. T. Tong, D. Barker / Tetrahedron Letters 47 (2006) 5017–5020
5019
and then Boc deprotection with TFA, neutralizing the
reaction mixture with added NEt3 to give (1R,10S)-1-
acetoxyquinolizidine 23 in 86% over three steps. Hydro-
lysis of the acetyl ester with NaOH gave (1R,10S)-9 in
80% yield, which was converted to (+)-(1S,10S)-epi-
quinamide 128 using the same procedure as outlined
for racemic 9, the only alteration in the synthesis being
the use of a higher yielding N-acetylation procedure.2
H
H
CO2H
CHO
(i)-(iii)
NH
NBoc
17
CH2OTBS
S
-3
18
(iv)
OTBS
OH
H
H
In summary, starting from commercially available pipe-
colinic acid 3, ( )-epiquinamide 1 was synthesized in
eight steps in 29% overall yield, whilst (+)-1 was synthe-
sized from S-3 in 12 steps in 13% overall yield. The key
step in the synthesis is the addition of acetylene 18 to
aldehyde 17 giving the desired erythro alkynol 19 in
good yield with high diastereoselectivity, with added
benefit that the undesired threo isomer cyclises to form
the easily separable oxazolidinone 20. The synthesis of
N-acyl analogues of epiquinamide is currently being
pursued, and their preparation and biological evaluation
will be reported in due course.
+
NBoc
19
O
N
20
OTBS
O
Scheme 4. Reagents, conditions and yields: (i) Boc2O, NEt3, dioxane,
H2O, rt, 18 h, 84%; (ii) (a) tBuCOCl, NEt3, DCM, ꢀ10 to 0 ꢁC, 90 min;
(b) (MeO)MeNHÆHCl, NEt3, DCM, 18 h, 72% over two steps; (iii)
LiAlH4, Et2O, 0 ꢁC to rt, 30 min, 99%; (iv) see Ref./note 26, 19 (69%)
and 20 (5%).
tion of the lithium anion of TBDMS-protected propar-
gyl alcohol 1822–24 to aldehyde 17 gave predominantly
(14:1, overall 79% yield) the desired erythro alkynol 19
and also the threo isomer 20, which was isolated in its
cyclised form (Scheme 4).25,26
Acknowledgements
The authors would like to thank The University of
Auckland, in particular the New Staff Research Fund
(#3604895/9346) for the financial support of this
project.
With alkyne 19 now having the carbons to make up the
quinolizidine ring system and the correct stereochemis-
try for transformation into 1 all that was required was
to close the second ring and convert the 1-hydroxyl
group into the required 1-acetamide. We initially inves-
tigated the idea of exchanging the hydroxyl group for an
azide at an earlier stage, thereby allowing us to hope-
fully reduce the azide and alkyne functionalities in 21
in a single synthetic step (Scheme 5). However, when
using our previously successful conditions, activation
of the hydroxyl group as a mesylate prior to azide addi-
tion resulted in an intramolecular attack of the Boc
group displacing the mesylate with resultant inversion
of stereochemistry to give the previously isolated oxazo-
lidinone 20.27 Alcohol 19 was protected as the acetyl es-
ter and the alkyne reduced to give silyl ether 22 in 77%
yield over two steps. Formation of the quinolizidine was
achieved by firstly deprotecting the silyl ether, using
TBAF, conversion of the free alcohol to the mesylate
References and notes
1. Fitch, R. W.; Garraffo, H. M.; Spande, T. F.; Yeh, H. J.
C.; Daly, J. W. J. Nat. Prod. 2003, 66, 1345–1350.
2. Wijdeven, M. A.; Botman, P. M. M.; Wijtmans, R.;
Schoemaker, H. E.; Rutjes, F. P. J. T.; Blaauw, R. H. Org.
Lett. 2005, 7, 4005–4007.
3. Huang, P.-Q.; Guo, Z.-Q.; Ruan, Y.-P. Org. Lett. 2006, 8,
1435–1438.
4. Sparatore, A.; Basilico, N.; Parapini, S.; Romeo, S.;
Novelli, F.; Sparatore, F.; Taramelli, D. Bioorg. Med.
Chem. 2005, 13, 5338–5345.
5. Collicutt, A. R.; Jones, G. J. Chem. Soc. 1960, 4104–4105.
6. Hadley, M. S.; King, F. D.; McRitche, B.; Turner, D. M.;
Wells, E. A. J. Med. Chem. 1985, 28, 1843–1847.
N3
H
OH
H
(i)
20
OTBS
NBoc
OTBS
NBoc
21 (not formed)
19
(ii), (iii)
NHAc
OAc
OAc
H
H
H
(vii)-(ix)
(iv)-(vi)
N
NBoc
22
OTBS
N
(+)-1
23
Scheme 5. Reagents, conditions and yields: (i) (a) MsCl, NEt3, DCM, 0 ꢁC, 30 min; (b) NaN3, 80 ꢁC, 18 h, 20 (64%); (ii) Ac2O, NEt3, DCM, rt, 4 h,
81%; (iii) H2, 10% Pd/C, MeOH, 2 h, 95%; (iv) TBAF, THF, rt, 3 h, 99%; (v) MsCl, NEt3, DCM, 0 ꢁC, 40 min, 92%; (vi) (a) TFA, DCM, 0 ꢁC,
90 min; (b) NEt3, DCM, 20 h, 95%; (vii) NaOH, EtOH, rt, 2 h, 80%; (viii) (a) 1.1 equiv MsCl, NEt3, DCM, 0 ꢁC, 30 min; (b) 5 equiv NaN3, 70–75 ꢁC,
18 h, 75% over two steps; (ix) (a) LiAlH4, THF, rt, 3 h; (b) Ac2O, 1 M NaOH, dioxane, rt, 3 h, 80% over two steps.

5020
S. T. Tong, D. Barker / Tetrahedron Letters 47 (2006) 5017–5020
7. Takehisa, K.; Kenji, K.; Shunichi, Y. Chem. Pharm. Bull.
afford the crude product, which was purified by flash
1967, 15, 337–344.
8. McFarlane, A. K.; Thomas, G.; Whiting, A. J. Chem.
Soc., Perkin Trans. 1 1995, 2803–2808.
chromatography (3:1 hexane–EtOAc) to provide 19
20
(234 mg, 69%) as a pale yellow oil: ½aꢂD ꢀ18.9 (c 0.5,
CHCl3);
m
max/cmꢀ1 3414 (O–H), 2930 (C–H), 2175
9. Leonard, N. J.; Swann, S., Jr.; Figueras, J., Jr. J. Am.
Chem. Soc. 1952, 74, 4620–4624.
10. Clemo, G. R.; Ramage, G. R. J. Chem. Soc. 1931, 437–
442.
11. Denmark, S. E.; Matsuhashi, H. J. Org. Chem. 2002, 67,
3479–3486.
12. Aaron, H. S.; Wicks, G. E.; Rader, C. P. J. Org. Chem.
1964, 29, 2248–2252.
13. Aaron, H. S.; Wicks, G. E.; Rader, C. P. J. Org. Chem.
1964, 29, 2252–2256.
14. Moehrle, H.; Karl, C.; Scheidegger, U. Tetrahedron 1968,
24, 6813–6824.
15. Hassner, A.; Maurya, R.; Padwa, A.; Bullock, W. H.
J. Org. Chem. 1991, 56, 2775–2781.
16. Chang, M.-Y.; Hsu, R.-T.; Tseng, T.-W.; Sun, P.-P.;
Chang, N.-C. Tetrahedron 2004, 60, 5545–5550.
17. Drag, M.; Lataj, R.; Gumienna-Kontecka, E.; Kozlowski,
H.; Kafarski, P. Tetrahedron: Asymmetry 2003, 14, 1837–
1845.
18. Aldehyde 13 was used immediately upon synthesis as it
was found to decompose in 1–2 days even if stored at low
temperatures.
19. Balboni, G.; Marastoni, M.; Merighi, S.; Borea; Pier, A.;
Tomatis, R. E. J. Med. Chem. 2000, 35, 979–988.
20. Thai, D. L.; Sapko, M. T.; Reiter, C. T.; Bierer, D. E.;
Perel, J. M. J. Med. Chem. 1998, 41, 591–601.
21. Portoghese, P. S.; Pazdernik, T. L.; Kuhn, W. L.; Hite, G.;
Shafier, A. J. Med. Chem. 1968, 11, 12–15.
22. Logue, M. W.; Teng, K. J. Org. Chem. 1982, 47, 2549–
2553.
(C„C), 1693 (C@O); dH (400 MHz, CDCl3; Me4Si) 0.10
(6H, s, Si(CH3)2), 0.90 (9H, s, SiMe2C(CH3)3), 1.47 (9H, s,
CO2C(CH3)3), 1.56–1.75 (5H, m, 3b-H, 4-H · 2, 5-H · 2),
1.97 (1H, br d, J = 13.2 Hz, 3a-H), 3.10–2.90 (1H, br s, 6a-
H), 3.95 (1H, br d, J = 10.8 Hz, 6b-H), 4.18 (1H, dd,
J1 = 10.1 Hz, J2 = 6.2 Hz, 2-H), 4.32 (2H, d, J = 1.7 Hz,
40-H), 4.65 (1H, dd, J1 = 6.3 Hz, J2 = 6.3 Hz, 10-H); dC
(100 MHz; CDCl3; Me4Si) ꢀ5.30 (Si(CH3)2), 18.11
(SiMe2C(CH3)3), 19.08 (C-4), 24.12 (C-3 or C-5), 24.57
(C-3 or C-5), 25.67 (SiMe2C(CH3)3), 28.28 (CO2C(CH3)3),
40.38 (C-6), 51.65 (C-40), 55.11 (C-2), 62.60 (C-10), 79.52
(CO2C(CH3)3), 83.86 (C-20 or C-30), 83.94 (C-20 or C-30),
155.39 (quat. C@O); m/z (FAB, NBA) 384 (0.14, M+H+),
328 (0.29), 310 (0.24), 284 (0.29), 184 (0.25,
MꢀCH(OH)C„CCH2OTBS), 128 (1.00, Mꢀ[CH-
(OH)C„CCH2OTBS+CðCH3Þþ]), 84 (0.62, N(CH2)5);
3
HRMS calculated for C20H38NO4Si (M+H+) 384.25701,
found 384.25768; and 20 (14 mg, 5%) as a pale yellow oil:
20
½aꢂD ꢀ18.9 (c 0.5, CHCl3); mmax/cmꢀ1 2929 (C–H), 2266
(C„C), 1758 (C@O), 1415, 1260; dH (300 MHz, CDCl3,
Me4Si) 0.12 (6H, s, Si(CH3)2), 0.91 (9H, s, C(CH3)3), 1.35–
1.44 (2H, m, 6a-H, 7a-H), 1.55–1.75 (2H, m, 8a-H, 6b-H),
1.74–1.84 (1H, m, 8b-H), 1.90–2.03 (1H, m, 7b-H), 2.83
(1H, td, J1 = 12.7 Hz, J2 = 3.4 Hz, 5a-H), 3.66 (1H, ddd,
J1 = 11.5 Hz, J2 = 8.0 Hz, J3 = 3.7 Hz, 8a-H), 3.88 (1H,
dd, J1 = 13.2 Hz, J2 = 4.4 Hz, 5b-H), 4.37 (2H, d,
J = 1.7 Hz, CH2O), 5.22 (1H, dt, J1 = 8.0 Hz,
J2 = 1.6 Hz, 1-H); dC (75.5 MHz, CDCl3, Me4Si) ꢀ5.22
(Si(CH3)2), 18.21 (C(CH3)3), 22.68 (C-4), 23.88 (C-3),
25.71 (C(CH3)3), 27.21 (C-5), 41.98 (C-2), 51.52 (CH2O),
56.93 (C-6), 67.42 (C-7), 88.77 (C„C), 156.0 (quat. C@O);
m/z (FAB, NBA) 310 (1.00, M+H+), 154 (0.28), 134
(0.40), 73 (0.40); HRMS calculated for C16H28NO3Si
(M+H+) 310.18385, found 310.18412.
23. Ridgway, B. H.; Woerpel, K. A. J. Org. Chem. 1998, 63,
458–460.
24. Gruza, H.; Kiciak, K.; Krasinski, A.; Jurczak, J. Tetra-
hedron: Asymmetry 1997, 8, 2627–2631.
25. syn-1-N-Boc-2-Alcohols of this form commonly cyclise to
form oxazolidinones, for examples, see: Beak, P.; Lee, W.
J. Org. Chem. 1990, 55, 2578–2580.
27. We later were to discover that if at any time before the Boc
group is removed and quinolizidine ring formation is
complete that this hydroxyl group was converted to a
mesylate then oxazolidinone formation is the predominant
reaction, thus necessitating the protection of the hydroxyl
group until that stage.
26. To a stirred solution of alkyne 18 (300 mg, 1.76 mmol) in
THF (5 mL) at ꢀ78 ꢁC under N2 was slowly added n-BuLi
(1.6 M, 1.05 mL, 1.68 mmol). The mixture was stirred at
0 ꢁC for 75 min and then cooled to ꢀ78 ꢁC. A solution of
aldehyde 17 (190 mg, 0.88 mmol) in THF (3 mL) was
added dropwise to the mixture and stirred at 0 ꢁC for 4 h.
After that, the mixture was left stirring at room temper-
ature overnight. The mixture was then diluted with 1 M
NaH2PO4 (20 mL) and extracted with EtOAc (20 mL · 3).
The combined organic extracts were washed with brine
(20 mL), dried (MgSO4) and concentrated in vacuo to
22
20
28. ½aꢂD +26.2 (c 0.25, CHCl3) [lit.2 ½aꢂD 28.0] (c 0.23, CHCl3);
dH (300 MHz; CDCl3; Me4Si) 1.22–1.72 (m, 9H), 1.79–
2.01 (m, 4H), 2.02 (s, 3H), 2.74–2.81 (m, 2H), 3.90–3.92
(m, 1H), 6.18 (br s, 1H); dC (75.5 MHz; CDCl3; Me4Si)
169.5, 64.2, 56.6, 56.5, 48.0, 29.6, 29.0, 25.5, 23.9, 23.5,
20.6; HRMS calculated for C11H21N2O (M+H+),
197.1654, found 197.1660. These values are consistent
with the literature values.1,2