Oxidative deamination of tetrahydroanabasine with o-quinones: An easy entry to lupinine, sparteine, and anabasine

Article abstract of DOI:10.1021/jo9602130

A mild oxidative deamination reaction of tetrahydroanabasine O-methyloxime 17 is described, making use of an o-quinone that is based on topaquinone (TPQ, 11), the cofactor that is present in copper-containing amine oxidases. In situ ring closure of the oxidation product produced double functionalized quinolizidine 5, containing an enamine functionality with excellent reactivity. From this quinolizidine 5 a variety of biogenetically related lupin alkaloids were prepared: lupinine (7) and aminolupinane (8) via reductive sequences and sparteine (9) via a condensation reaction with dehydropiperidine 1. The configurationally more favorable trans isomers epilupinine (25) and β-isosparteine (10) were formed when more drastic reaction conditions were used for oxime hydrolysis. Anabasine (4) and a new 5-piperidylanabasine derivative 6 were formed by an unexpected acid catalyzed ring transformation reaction, whereby the pyridine ring was formed via oxime-induced aromatization. The stereochemistry of the reaction products and the biogenetic implications are discussed.

Full text of DOI:10.1021/jo9602130

J . Org. Chem. 1996, 61, 5581-5586
5581
Oxid a tive Dea m in a tion of Tetr a h yd r oa n a ba sin e w ith o-Qu in on es:
An Ea sy En tr y to Lu p in in e, Sp a r tein e, a n d An a ba sin e
Martin J . Wanner and Gerrit-J an Koomen*
Amsterdam Institute of Molecular Studies, Laboratory of Organic Chemistry, University of Amsterdam,
Nieuwe Achtergracht 129, 1018 WS Amsterdam, The Netherlands
Received February 1, 1996X
A mild oxidative deamination reaction of tetrahydroanabasine O-methyloxime 17 is described,
making use of an o-quinone that is based on topaquinone (TPQ, 11), the cofactor that is present in
copper-containing amine oxidases. In situ ring closure of the oxidation product produced double-
functionalized quinolizidine 5, containing an enamine functionality with excellent reactivity. From
this quinolizidine 5 a variety of biogenetically related lupin alkaloids were prepared: lupinine (7)
and aminolupinane (8) via reductive sequences and sparteine (9) via a condensation reaction with
dehydropiperidine 1. The configurationally more favorable trans isomers epilupinine (25) and
â-isosparteine (10) were formed when more drastic reaction conditions were used for oxime
hydrolysis. Anabasine (4) and a new 5-piperidylanabasine derivative 6 were formed by an
unexpected acid catalyzed ring transformation reaction, whereby the pyridine ring was formed via
oxime-induced aromatization. The stereochemistry of the reaction products and the biogenetic
implications are discussed.
Sch em e 1
The lupin alkaloids have been an attractive synthetic
target for several decades. During this period an in-
creasing number of lupin alkaloids were identified from
several lupin species,1 and attention was given to their
biogenesis.1-3 Numerous labelling experiments have
been performed, leading to the conclusion that cadaver-
ine, and thus lysine, is the source of the carbon and
nitrogen atoms present in these alkaloids. Scho¨pf stud-
ied the chemistry of dehydropiperidine (1), the expected
product of the oxidative deamination of cadaverine in vivo
(Scheme 1). This imine is extremely reactive and dimer-
izes spontaneously in water at neutral pH to form
tetrahydroanabasine (2),4 which can be isolated as its
crystalline di-HBr salt.5 Solutions of both dehydropip-
eridine and tetrahydroanabasine are too unstable to be
obtained as such from plant material, which leaves the
exact intermediates of the biosynthesis unclear. Diamine
oxidase (DAO) catalyzed oxidative deamination of tet-
rahydroanabasine probably proceeds via the ring-opened,
primary amino compound and transforms the dipiperi-
dine ring system in 2 into quinolizidine 3, which contains
a characteristic structural unit of the lupin alkaloids.
This hypothetical, difunctionalized intermediate in the
biosynthesis of sparteine-, ormosanine-, and matrine-type
alkaloids is obtained from plants in its stable reduced
form: lupinine (7) and its isomer epilupinine (25).
This suggested bioroute seems plausible and empha-
sizes the possibilities of difunctionalized quinolizidine 5
as a useful synthetic intermediate for the laboratory
synthesis of the lupin alkaloids. We will describe here
the conversion of tetrahydroanabasine (2) into this reac-
tive quinolizidine 5 via a mild, o-quinone-catalyzed
oxidative deamination reaction that is related to the
biosynthetic process catalyzed by the topaquinone (TPQ)-
containing amine oxidases, the deaminating enzymes for,
e.g., cadaverine and putrescine in plants (eq 1).
X Abstract published in Advance ACS Abstracts, J uly 15, 1996.
(1) For a recent review see: Saito, K.; Murakoshi, I. Chemistry,
Biochemistry and Chemotaxonomy of Lupine Alkaloids in the Legu-
minosae. In Studies in Natural Products Chemistry 15; Atta-ur-
Rahman, Ed.; Elsevier: Amsterdam, 1995; p 519.
(2) (a) Golebiewski, W. M.; Spenser, I. D. Can. J . Chem. 1985, 63,
2707. (b) Brown, A. M.; Rycroft, D. S.; Robins, D. J . J . Chem. Soc.,
Perkin Trans. 1 1991, 2353. (c) Gavin, S. S.; Equi, A. M., Robins, D. J .
Can. J . Chem. 1994, 72, 31.
(3) Golebiewski, W. M.; Spenser, I. D. Can. J . Chem. 1988, 66, 1734.
(4) Schu¨tte, H. R. Arch. Pharm. (Weinheim) 1960, 293, 1006.
(5) Tetrahydroanabasine 2 can be prepared from both R- and iso-
tripiperidine: Scho¨pf, C.; Braun, F.; Koop, H.; Werner, G. Liebigs Ann.
Chem. 1962, 658, 156.
Conversion of this intermediate (5) into the natural
products anabasine, lupinine, sparteine, and â-iso-
sparteine will be described (Scheme 2). In addition to
S0022-3263(96)00213-7 CCC: $12.00 © 1996 American Chemical Society

5582 J . Org. Chem., Vol. 61, No. 16, 1996
Wanner and Koomen
Sch em e 2
Sch em e 3
these alkaloids, lupinamine and a new, 3-piperidyl-
substituted anabasine derivative are prepared from 5.
Since we use racemic tetrahydroanabasine, all of the
synthetic products are obtained as racemates; it should
be noted that often both antipodes and sometimes race-
mates of the natural alkaloids are isolated from different
plants. Non enzymic biosynthesis of these compounds
can explain this observation.6 In a later stage of develop-
ment, enzyme-catalyzed metabolism of one of the enan-
tiomers might lead to optically active secondary metabo-
lites.
Ch em istr y
Our synthetic approach starts with tetrahydroana-
basine (2), which can easily be prepared as its stable
trans isomer starting from dehydropiperidine trimer as
described by Scho¨pf.5 To obtain a free amino group,
which is required for oxidative deamination, the cyclic
imine in 2 was converted into an O-methyloxime (17,
Scheme 3). Although the oxime was isolated as a syn/
anti mixture, the stereochemistry around the C-C bond
was still preserved. In view of the limited stability of
the deaminated product 5, a mild oxidative procedure for
this primary amine was required, and only the com-
mercially available o-quinone 13 seemed suitable for this
purpose (Scheme 4). Two tert-butyl substituents in 13
prevent conjugate addition reactions of nucleophiles to
the reactive double bonds. It should be noted that this
reagent was used for deamination reactions by Corey and
Sch em e 4
(6) McLean, S.; Misra, R.; Kumar, V.; Lamberton, J . A. Can. J .
Chem. 1981, 59, 34.
(7) Corey E. J .; Achiwa, K. J . Am. Chem. Soc. 1969, 91, 1429. For
further studies with quinone 13, see: Klein, R. F. X.; Bargas, L. M.;
Horak, V. J . Org. Chem. 1988, 53, 5994. Vander Zwan, M. C.; Hartner,
F. W.; Reamer, R. A.; Tull, R. J . Org. Chem. 1978, 43, 509.
Achiwa7 in 1969, several years before the comparable
TPQ (11, eq 1) was found as the active cofactor in copper-
containing amine oxidases.8 Recently, several TPQ

Deamination of Tetrahydroanabasine with o-Quinones
J . Org. Chem., Vol. 61, No. 16, 1996 5583
models were developed9 for the oxidized form of the 2,4,5-
trihydroxyphenylalanine (TOPA) cofactor, allowing stud-
ies toward the enzyme reaction. Until now, a syntheti-
cally useful catalyst that is based on TPQ has not been
available since numerous side reactions occur in this
complicated redox system. The mechanism of oxidation
is shown in Scheme 4, where aromatization of 14 to 15
is the driving force of the reaction. According to Corey
and Achiwa7 ketones are obtained in high yield from 13
with primary, R-branched amines such as cyclohexyl-
amine. Primary amines containing an R-unbranched
substituent (e.g., benzylamine) are not suitable as sub-
strate in the oxidative deamination with 13 since via a
second quinone-catalyzed oxidation step benzoxazoles
(16) are formed instead of aldehydes. In our substrate
(17) overoxidation to the corresponding benzoxazole is
prevented by the presence of a second amino group, which
takes over the imino carbon atom initially formed and
leads to a 6-membered cyclic iminium salt (18, Scheme
3). Thus, alkaline workup of the reaction of 17 with 13
provides quinolizidine 5 (45%), a reactive enamine that
is our central precursor for the synthesis of several lupin
alkaloids. Again, the stereochemistry was preserved,
providing a cis relationship around the quinolizidine C1-
C10 bond. 5 is stable for longer periods at -20 °C or as
its hydrochloride in solution at room temperature. In
contrast, enzymic oxidative deamination of 17 using
commercially available porcine kidney diamine oxidase
was not successful due to slow conversion of the sub-
strate. Especially at neutral pH, which is required for
this enzymic process, undesired reactions with the prod-
uct 5 take place easily. The instability of 5 is demon-
strated by stirring a solution of 5 overnight in aqueous
methanol at pH 7, yielding 80% of dimer 19 as a result
of enamine/imine equilibration followed by an iminoaldol
condensation between 18 and 5 (Scheme 3). This dimer
was present as a complex mixture of isomers at C4′,
whereby each isomer consisted of four syn/anti isomers
of the oximes.
Sch em e 5
Sch em e 6
Lu p in a m in e.12 Reduction of 5 with NaBH4 in metha-
nol gave a stable quinolizidine 22, which was further
reduced with LiAlH4 to give lupinamine 8, characterized
as its N-acetyl derivative 23 (Scheme 6).
An a ba sin e.13 Several methods for cleaving O-alkyl
oximes are available.14 Simple hydrochloric acid-cata-
lyzed hydrolysis of 5 in acetone/water mixtures, however,
gave an unexpected rearrangement, whereby oxime-
induced aromatization of the piperidine ring gave racemic
anabasine (4) in one step (Scheme 6). It should be noted
that anabasine obtained from Nicotiana species is bio-
genetically derived from dehydropiperidine and nicotinic
acid13 and not from two molecules of dehydropiperidine.
Although in vitro formation of anabasine is demonstrated
with pea and lupine extracts15 this has not been con-
firmed by in vivo labelling experiments.
Lu p in in e.16 Milder hydrolytic conditions for oxime
hydrolysis are based on reduction of the N-O bond,
resulting in the formation of an imine, which is vulner-
able to hydrolytic conditions. Aqueous sodium hydrogen
sulfite, sodium dithionite, or TiCl3/HCl indeed hydrolyzed
the oxime at elevated temperatures; however, as a result
of enolization of the intermediate aldehyde 24 (Scheme
7) the alcohols that were obtained in low yields after
NaBH4 workup consisted of mixtures of lupinine (7) and
Direct oxidation of tetrahydroanabasine 2 was at-
tempted with quinone 13 in the presence of water,
assuming that in an equilibrium the imine in 2 is
hydrolyzed to its ring-opened amino aldehyde form
(Scheme 1). Reaction of 2 with o-quinone 13 in a sodium
acetate-buffered solution produced a crystalline 1:1 ad-
dition product, according to HRMS and NMR data. X-ray
analysis10 was required to establish its structure as 21:
the 4 + 2 cycloaddition product derived from the enamine
20 of tetrahydroanabasine (Scheme 5). Comparable
examples of cycloaddition reactions between less substi-
tuted enamines and o-quinones have been described in
the literature.11
(8) (a) J anes, S. M.; Mu, S.; Wemmer, D., Smith, A. J .; Kaur, S.;
Maltby, D.; Burlingame, A. L.; Klinman, J . P. Science 1990, 248, 981.
(b) J anes, S. M.; Palcic, M. M.; Scaman, C. H., Smith, A. J .; Brown, D.
E.; Dooley, D. M.; Mure, M.; Klinman, J . P. Biochemistry 1992, 31,
12147. (c) Dooley, D. M.; McGuirl, M. A.; Brown, D. E.; Turowski, P.
N.; McIntire, W. S.; Knowles, P. F. Nature 1991, 349, 262.
(9) Mure, M.; Klinman, J . P. J . Am. Chem. Soc. 1995, 117, 8698
and 8707. Lee, Y.; Sayre, L. M. J . Am. Chem. Soc. 1995, 117, 11823.
(10) The author has deposited atomic coordinates for this structure
with the Cambridge Crystallographic Data Centre. The coordinates
can be obtained, on request, from the Director, Cambridge Crystal-
lographic Data Centre, 12 Union Road, Cambridge, CB2 1EZ, UK. We
wish to thank J . Fraanje, K. Goubitz, and H. Schenk of the Department
of Crystallography of this University for the X-ray crystal structure
determination of compound 21.
(13) (a) Watson, A. B.; Brown, A. M.; Colquhoun, I. J .; Walton, N.
J .; Robins, D. J . J . Chem. Soc., Perkin Trans. 1 1990, 2607. (b) Gross,
D. In Biochemistry of Alkaloids; Mothes, K., Schu¨tte, H. R., Luckner,
M., Eds.; VCH publishers: Deerfield Beach, 1985; p 163.
(14) Weitz, D. J .; Bednarski, M. D. J . Org. Chem. 1989, 54, 4957
and references cited therein.
(11) Ried, W.; Torok, E. Liebigs Ann. Chem. 1965, 687, 187.
(12) Recent syntheses of lupinamine: (a) Wanner, M. J .; Koomen
G.-J . Tetrahedron 1991, 47, 8431. (b) Michael, J . P.; J ungmann, C. M.
Tetrahedron 1992, 48, 10211.
(15) Mothes, K.; Schu¨tte, H. R.; Simon, H.; Weygand, F. Z. Natur-
forsch., Teil B 1959, 14, 49.
(16) For a recent synthesis of lupinine and epilupinine, see: Hua,
D. H.; Miao, S. W.; Bravo, A. A.; Takemoto, D. J . Synthesis 1991, 970.

5584 J . Org. Chem., Vol. 61, No. 16, 1996
Wanner and Koomen
Sch em e 7
Sch em e 9
Sch em e 8
indicate the presence of only one diastereoisomer18 the
possibility that 6 consists of two diastereoisomers cannot
be excluded.
Sp a r tein e19 a n d â-Isosp a r tein e.20 Removal of the
enamine double bond in 26 should prevent the afore-
mentioned rearrangement into 5-piperidylanabasine dur-
ing acid-catalyzed oxime hydrolysis. Acid free hydroge-
nation with H2/Pd was not suitable, however, since
reduction of the oxime was faster then reduction of the
enamine. Reduction with NaCNBH3 and acetic acid
(Scheme 8) established the stereochemistry at C3 via
protonation of the enamine, providing 27. After reduc-
tion 28 was obtained with a trans relationship between
C1 and C3, which is unsuitable for cyclization. We
returned to compound 26, which was subjected to oxida-
tive removal of the oxime (Scheme 9) using ozone and
acid, as described for the synthesis of lupinine in Scheme
7. The reaction proceeded very slowly and incomplete
this time, probably because ozone has to react with a
species that is protonated at two nitrogen atoms. Long
reaction times and higher temperatures can give rise to
several side reactions, according to a recent publication
describing the ozonolysis of O-methyl oximes at 0 °C.21
During the ring closure of 29 to 30 with NaOAc/HOAc
an imine/enamine equilibrium takes place, enabling the
formation of an intermediate with the substituents at C1
and C3 in a 1,3 diaxial position, which is required for
cyclization. After NaCNBH3 reduction of the 1,10;16,-
17-didehydrosparteinium ion 30, formed in a NaOAc/
HOAc buffer, sparteine was obtained as the only isomer
(21%). Reductive hydrolysis of the oxime functionality
in 26 with TiCl3/HCl at high temperatures (80-90 °C)
was also attempted. Reduction of the intermediate
iminium salts 30 with NaCNBH3 followed by chroma-
tography yielded â-isosparteine (10, 11%) and sparteine
(9, 9%).
the thermodynamically more stable epilupinine (25). An
efficient oxidative removal of O-methyl oximes consists
of treatment with ozone in methanol.14 Trifluoroacetic
acid was used to protonate the tertiary amine and to
protect it from oxidation. Ozonolysis at -50 °C followed
by the addition of NaBH4 at low temperature yielded
crystalline lupinine (7) as one isomer.
5-P ip er id yla n a b a sin e. Sparteine and isomers are
built up from three dehydropiperidine equivalents. Ef-
ficient condensation of 5 with dehydropiperidine (1) took
place when optimized conditions were used (Scheme 8).
Without precautions dimerization of both starting ma-
terials occurred to form 19 and tetrahydroanabasine (2).
In situ monomerized R-tripiperidine17 in HOAc/NaOAc
buffered methanol gave 80% 3-piperidylquinolizidine 26
after a few hours at room temperature. This product was
present as a mixture of isomers at C12 (each isomer as
a syn/anti mixture of oximes), which was difficult to
identify because in the 1H and 13C NMR spectra only
small chemical shift differences were observed. During
the condensation reaction the stereochemistry at C1 and
C10 was not affected. Enamine 26 should be directly
suitable for cyclization to sparteine after hydrolysis of
the oxime. Heating 26 in aqueous HCl again gave the
same rearrangement/aromatization process as observed
for anabasine (4, Scheme 6) resulting in the formation
of 5-piperidylanabasine (6) in moderate yield. The ster-
eochemical relationship between the two piperidine rings
could not be established because of the large distance
between the asymmetric centers. Although NMR spectra
In view of the low yields, mechanistic statements
concerning the stereochemical course of these reactions
cannot be made. Considering the mildness of the ozo-
nolysis with respect to isomerization, it is acceptable that
during the synthesis of sparteine itself, the original
stereocenters remain intact. The fourth, newly formed
asymmetric carbon atom 9 will always adopt a cis
relation with the other bridgehead carbon atom 7. The
more symmetric isomer â-isosparteine is thermodynami-
cally favorable and is formed as a result of one or more
isomerization processes. The third possible isomer,
R-isosparteine, was not observed in these reactions.
(17) The corresponding symmetric trimer of 1 is an easy to handle,
crystalline substance. Dissolving this trimer in dilute HCl results in
a rapid hydrolysis to the protonated monomeric form of 1: (a) Scho¨pf,
C.; Braun, F.; Komzak, A. Chem. Ber. 1956, 89, 1821. (b) Wanner, M.
J .; Velzel, A. W.; Koomen, G.-J . J . Chem. Soc., Chem. Commun. 1993,
174.
(19) Bohlmann, F.; Zeisberg, R. Chem. Ber. 1975, 108, 1043.
(20) Bohlmann, F.; Schumann, D.; Arndt, C. Tetrahedron Lett. 1965,
2703.
(21) Griesbaum, K.; Ovez, B.; Sung Huh, T.; Dong, Y. Liebigs Ann.
Chem. 1995, 1571.
(18) In 13C NMR only C3 of the piperidine ring was double (∆δ )
0.08 ppm).

Deamination of Tetrahydroanabasine with o-Quinones
J . Org. Chem., Vol. 61, No. 16, 1996 5585
31.6, 29.8, 27.6, 25.6, 25.3, 25.1, 22.2; IR (CHCl3) 1586, 1482
cm-1; HRMS obsd mass 387.3012, calcd for C24H39N2O2 (M +
1) 387.3011.
Further improvement of the oxime hydrolysis proce-
dures is necessary to study the stereochemical processes
during sparteine formation.
Qu in olizid in e 22. Quinolizidine 5 (0.194 g, 1 mmol) was
dissolved in MeOH (5 mL) and immediately reduced with
NaBH4 (0.076 g, 2 mmol) for 3 h at rt. The reaction mixture
was concentrated in vacuum, quenched with 1 M HCl, and
extracted with ether after addition of aqueous K2CO3. After
chromatography (CH2Cl2/MeOH/concd NH4OH 90/10/1) 22
(0.174 g, 89%) was obtained as a syrup: syn:anti ) 1:4; 1H
NMR (CDCl3) δ 7.78 (d, J ) 8.8 Hz, 1H), 7.10 (d, J ) 7.5 Hz,
1H), 3.84 (s, 3H), 3.81 (s, 3H), 3.0-1.15 (m, 16H); 13C NMR
(CDCl3) δ 153.3 (syn), 152.8 (anti), 64.1, 63.8, 61.4, 61.2, 57.0,
56.9, 39.9, 36.0, 30.9, 30.6, 20.0, 25.7, 25.6, 24.6, 24.6, 22.4,
Exp er im en ta l Section 22
Tetr a h yd r oa n a ba sin e Oxim e (17). Crystalline 2 (di-HBr
salt, hydrate) was prepared from dehydropiperidine trimer 1
as described by Scho¨pf et al.5 2 (di-HBr salt, hydrate, 3.46 g,
10 mmol) was stirred overnight at rt with methoxyamine (4.5
mL of a 35% solution in water, 30 mmol) in methanol (35 mL).
Water (ca. 5 mL) was added, followed by an excess amount of
solid K2CO3. After the solution was stirred for 5 min, ether
was added followed by enough Na2SO4 to bind the water that
was present. The mixture was stirred for 1 h and filtrated,
and the salts were washed several times with dry ether. The
solvents were removed, and the resulting liquid was kept on
the evaporator (45 °C, 20 mbar) for 2 h, yielding almost pure
17 (2.08 g, 98.5%) as an oil: syn:anti ) 1:4; 1H NMR (CDCl3)
δ 7.23 (d, J ) 8.6 Hz, 1H, anti oxime), 6.52 (d, J ) 8.4 Hz, 1H,
syn oxime), 3.80 (s, 3H), 3.79 (s, 3H), 3.03 (m, 1H), 2.75-2.5
(m, 4H), 2.15 (m, 1H), 1.9-1.1 (m, 10 H); 13C NMR (CDCl3) δ
153.1 (syn oxime), 152.2 (anti oxime), 61.2, 58.6, 47.1, 45.3,
41.9, 31.2, 30.3, 26.4, 26.2, 24.7; IR (CHCl3) 1585 cm-1; HRMS
obsd mass 214.1929, calcd for C11H24N3O (M + 1) 214.1883.
Qu in olizid in e 5. A solution of quinone 13 (1.10 g, 5 mmol)
in THF (10 mL) was added dropwise to a stirred solution of
oxime 17 (0.852 g, 4 mmol) in MeOH (35 mL) at 0 °C in 5
min. After 10 min at 0 °C, the dark reaction mixture was
stirred at rt for 15 min and quickly concentrated to a small
volume in vacuum (bath T < 30 °C). The residue was diluted
with 5% HCl solution (100 mL) and ether (100 mL). The
organic layer was washed twice with 5% HCl solution, and
the combined aqueous layers were washed with ether. Solid
K2CO3 was added, and enamine 5 was extracted with ether.
In general, better yields were obtained when 5 was not purified
but immediately used for further reactions. Flash chroma-
tography (CH2Cl2/MeOH/concd NH4OH 90/10/1) gave 5 (0.35
g, 45%) as a syrup: syn:anti ) 1:2; 1H NMR (CDCl3) δ 7.47 (d,
J ) 8.5 Hz, 1H), 6.85 (d, J ) 7.8 Hz, 1H), 5.72 (dd, J ) 8 Hz,
1.8 Hz, 1H), 4.42 (m, 1H), 3.85 (s, 3H), 3.82 (s, 3H), 3.0-3.05
(m, 1H), 2.9-1.1 (m, 11H); 13C NMR (CDCl3) δ 152.7, 152.6,
137.5, 137.3, 96.3, 95.7, 61.4, 61.1, 58.0, 57.6, 53.0, 38.4, 31.6,
29.5, 29.0, 26.1, 25.9, 25.7, 25.6, 24.4.
21.6; IR (CHCl3) 2800, 2760, 1650 cm-1
196.1578, calcd for C24H39N2O2 196.1565.
. HRMS obsd mass
N-Acetyllu p in a m in e (8). Oxime 22 (0.050 g, 0.25 mmol)
was reduced with LAH (0.076 g, 2 mmol) in refluxing THF (2
mL) for 7 h. Workup with NaOH and ether gave the free
amine as an oil, which was acetylated with a mixture of ethyl
acetate and acetic anhydride (reflux, 1 h). The solvents were
evaporated, and the residue was treated with aqueous am-
monia. Ether extraction and recrystallization of the crude
product (PE/EtOAc) gave 812 (45 mg, 85%): mp 121-122 °C
(needles); 1H NMR (CDCl3) δ 7.45 (br, 1H), 3.49 (dddd, J )
3.0, 5.7, 8.7, 13.9 Hz, 1H), 3.38 (dddd, J ) 3.5, 6.7, 10.3, 13.9
Hz, 1H), 2.84 (m, 2H), 1.96 (s, 3H), 2.1-1.45 (m, 13H), 1.2-
1.25 (m, 1H); 13C NMR (CDCl3) δ 169.7, 64.8, 57.3, 57.1, 41.4,
36.0, 30.0, 29.8, 25.7, 24.8, 23.5, 22.0; IR (CHCl3) 3460, 2800,
2760, 1650, 1520 cm-1; HRMS obsd mass 211.1790, calcd for
C12H23N2O (M + 1) 211.1884.
(()-An a ba sin e (4). Quinolizidine 5 (0.039 g, 0.2 mmol) in
aqueous HCl (5%, 0.5 mL) was kept at 90 °C for 24 h. The
reaction mixture was made alkaline with excess solid K2CO3,
and the product was extracted with ether. Chromatography
(CH2Cl2/MeOH/concd NH4OH 90/10/1) gave 413 (0.018 g, 56%)
as an oil: 1H NMR (CDCl3) δ 8.56 (d, J ) 1.7 Hz, 1H), 8.46
(dd, J ) 4.7, 1.4 Hz, 1H), 7.70 (ddd, J ) 7.9, 1.8, 1.7 Hz, 1H),
7.22 (dd, J ) 4.7, 7.9 Hz, 1H), 3.61 (dd, J ) 10.5, 2.6 Hz, 1H),
3.18 (bd, J ) 11.4 Hz, 1H), 2.78 (m, 1H), 2.0-1.4 (m, 6H); 13
C
NMR (CDCl3) δ 148.7, 148.6, 140.6, 134.1, 123.4, 59.8, 47.6,
34.8, 25.7, 25.2; IR (CHCl3) 1590, 1500 cm-1; HRMS obsd mass
162.1161, calcd for C10H14N2 162.1138.
Ozon olysis of 22: (()-Lu p in in e (7). TFA (23 µL, 0.3
mmol) was added to a solution of oxime 22 (0.047 g, 0.24 mmol)
in MeOH (3 mL). The solution was cooled to -60 °C and
treated with ozone in portions. The bath temperature was
kept between -50 and -60 °C, and after 90 min the solution
was purged with nitrogen followed by the addition of NaBH4
(0.038 g, 1 mmol). The reaction mixture was stirred at rt for
1 h, the solvent was evaporated, and the residue was treated
with NaOH solution (1M, 5 mL). Addition of solid K2CO3 and
ether extraction yielded pure lupinine16 (7) (0.022 g, 54%): mp
Qu in olizid in e Dim er 19. A solution of 5 (0.049 g, 0.25
mmol) and acetic acid (15 µL, 0.025 mmol) in MeOH (1 mL)
was stirred at rt for 20 h. Workup with aqueous K2CO3 and
ether yielded after chromatography (CH2Cl2/MeOH/concd NH4-
OH 95/5/0.5) dimer 19 (39 mg, 80%) as a mixture of isomers:
1H NMR (CDCl3) δ 7.9-6.5 (m, 10 different CH)N signals),
5.67 (broad, 1H, NCH)C), 3.86-3.80 (10 CH3 signals); IR
(CHCl3) 1585 cm-1; HRMS obsd mass 388.2847, calcd for
C22H36N4O2 388.2835.
1
58-59 °C (ether/hexane); H NMR (CDCl3) δ 4.14 (ddd, J )
Cycloa d d ition of Tetr a h yd r oa n a ba sin e to 3,5-Di-ter t-
bu tyl-1,2-ben zoqu in on e (21). Crystalline 2 (di-HBr salt,
hydrate, 0.178 g, 0.5 mmol) was dissolved in MeOH (6 mL)
together with NaOAc (0.082 g, 1 mmol). A solution of 13 (0.121
g, 0.55 mmol) in THF (1 mL) was added in one portion. The
dark reaction mixture was stirred overnight and filtrated,
yielding 2110 (0.085 g, 44%) as a white solid. When no
crystallization takes place, alkaline workup (K2CO3, H2O,
ether) and trituration with MeOH gives comparable results.
The filtrate contained small amounts of an isomer that could
not be obtained in pure form. 21: mp 214-218 °C (from
MeOH) and 225-228 °C (from CH2Cl2/PE); IR (CHCl3) 1586,
10.7, 5.6, 1.2 Hz, 1H), 3.68 (d, J ) 10.8 Hz, 1H), 2.81 (m, 2H),
2.15 (m, 2H), 2.01 (m, 1H), 1.9-1.45 (m, 10H), 11.25 (m, 1H);
13C NMR (CDCl3) δ 66.0, 65.1, 57.1, 57.0, 38.1, 31.4, 29.7, 25.6,
24.6, 22.9; IR (CHCl3) 3000-3500, 2810, 2770 cm-1; HRMS
obsd mass 169.1460, calcd for C10H19NO 169.1501.
TiCl3 Red u ction of 22: (()-Lu p in in e (7) a n d (()-
Ep ilu p in in e (25). A solution of 22 (0.047 g, 0.24 mmol) in
aqueous TiCl3 (10% solution in 25% HCl, 1 mL) was adjusted
to pH 3-4 by the addition of Na2CO3. The resulting suspen-
sion was heated at 50 °C for one night, and the products were
extracted with ether after basification with K2CO3. According
1
to H NMR the obtained product (12 mg, 30%) consisted of a
1
1482 cm-1; H NMR (CDCl3) δ 6.83 (d, J ) 2.3 Hz, 1H), 6.74
5:2 mixture of lupinine and epilupinine.16 Epilupinine: 1H
NMR (CDCl3) δ 3.64 (dd, J ) 3.7, 10.8 Hz, 1H), 3.55 (dd, J )
10.8, 5.6 Hz, 1H).
(d, J ) 2.3 Hz, 1H), 5.19 (d, J ) 0.5 Hz, 1H), 3.28 (m, J )
11.1, 2.7 Hz, 1H), 3.18 (ddd, J ) 10.8, 10.8, 4.1 Hz, 1H), 3.10
(m, J ) 13 Hz, 1H), 2.76 (m, J ) 11.1, 4.1, 4.1 Hz, 1H), 2.62
(ddd, J ) 13, 12.5, 2.8 Hz, 1H), 1.96 (m, 1H), 1.90-1.2 (m,
9H), 1.39 (s, 9H), 1.28 (m, 9H); 13C NMR (CDCl3) δ 142.6, 141.0,
138.2, 137.3, 115.2, 112.0, 81.4, 75.3, 57.1, 46.9, 39.2, 35.0, 34.3,
3-P ip er id ylqu in olizid in e 26. R-Tripiperidine5,15 (1) (0.332
g, 4 mmol) was stirred with aqueous HCl (30%, 1 mL) until a
clear solution of 2 (hydrochloride) was obtained (ca. 1 h).
Quinolizidine 5 (0.582 g, 3 mmol) was dissolved in MeOH (15
mL) and immediately mixed with the R-tripiperidine monomer
(1) and sodium acetate (0.82 g, 10 mmol). After the mixture
was stirred at rt for 3 h the MeOH was evaporated and the
(22) For general information, see: Wanner, M. J .; Koomen G.-J . J .
Org. Chem. 1994, 59, 7479.

5586 J . Org. Chem., Vol. 61, No. 16, 1996
Wanner and Koomen
residue was diluted with K2CO3 solution and extracted three
times with ether. Additional solid K2CO3 may be added to the
water layer to improve the extraction. Chromatography (CH2-
Cl2/MeOH/concd NH4OH 90/10/1) gave 26 (0.664 g, 80%) as a
1:1 mixture of isomers, each as a 1:2 mixture of syn and anti
oximes: 1H NMR (CDCl3) δ 7.370 (d, J ) 8.4 Hz, 1H), 7.366
(d, J ) 8.4 Hz, 1H), 6.751 (d, J ) 7.8 Hz, 1H), 6.749 (d, J )
7.8 Hz, 1H), 3.84, 3.83, 3.81, 3.79 (4 × CH3), 3.3-1.2 (m, 23H);
13C NMR (CDCl3) δ 152.4, 152.4, 152.3, 152.2 (4 × CH)NO)
133.1, 132.8, 132.7, 132.5, (4 × NCH)), 112.1, 111.9, 11.6,
111.4 (4 × C-quart.); IR (CHCl3) 1585 cm-1; HRMS obsd mass
277.2172, calcd for C16H27N3O 277.2108.
(()-5-P ip er id yla n a ba sin e (6). A solution of 26 (0.028 g,
0.1 mmol) in aqueous HCl (5%, 2 mL) was kept at 80 °C for 5
h. The reaction mixture was cooled, made alkaline with K2-
CO3, and extracted with ether. Chromatography (CH2Cl2/
MeOH/concd NH4OH 85/15/1.5) gave 5 (0.008 g, 35%) as a
glass: 1H NMR (CDCl3) δ 8.45 (d, J ) 1.9 Hz, 2H), 7.72 (bs,
1H), 3.62 (dd, J ) 10.2, 2.4 Hz, 2H), 3.18 (bd, J ) 11.5 Hz, 2
H), 2.78 (ddd, J ) 11.5, 11.5, 2.7 Hz, 2H),1.95-1.4 (m, 12H);
13C NMR (CDCl3) δ 147.4, 140.4, 132.4, 59.9, 47.7, 34.9, 34.8,
25.8, 25.3; IR (CHCl3) 1595 cm-1; HRMS obsd mass 245.1879,
calcd for C15H23N3 245.1825.
rt for 10 min, sodium acetate (0.492 g, 6 mmol) and acetic acid
(2 mL) were added and stirring was continued for 1 h.
Reduction was performed by adding NaCNBH3 (0.126 g, 2
mmol) and stirring overnight. Concd HCl (2 mL) was added,
and after being stirred for 15 min the reaction mixture was
concentrated in vacuum, diluted with water, and made alka-
line with excess K2CO3. Ether extraction and chromatography
(CH2Cl2/MeOH/concd NH4OH 85/15/1.5) gave sparteine (9)
(0.025 g, 21%) as an oil: 1H NMR (CDCl3) δ 2.84 (bd, J ) 11.4
Hz, 1H), 2.78 (dd, J ) 10.9, 10.8 Hz, 1H), 2.69 (bd, J ) 10.8
Hz, 1H), 2.54 (bd, J ) 10.9 Hz, 1H), 2.38 (dd, J ) 11.3, 3.5
Hz, 1H), 2.10 (m, 2H), 2.00 (dd, J ) 11.0, 2.6 Hz, 1H), 1.95
(m, 1H), 1.85 (m, 1H), 1.75-1.15 (m, 15H), 1.08 (ddd, J ) 12.1,
2.5, 2.5 Hz, 1H); 13C NMR (CDCl3) δ 66.4, 64.4, 61.8, 56.2, 55.3,
53.2, 35.9, 34.0, 30.0, 29.3, 27.5, 25.8, 25.4, 24.7, 24.6; HRMS
obsd mass 234.2079, calcd for C15H26N2 234.2096.
(()-â-Isosp a r tein e20 (10). TiCl3 (0.308 g, 2 mmol) was
added to a solution of oxime 26 (0.100 g, 0.36 mmol) in 5%
HCl (4 mL), and the resulting mixture was heated at 80 °C
overnight. After the mixture was cooled to rt, water (5 mL)
was added and the HCl was neutralized with NaOAc until the
mixture was slightly acidic. Additional acetic acid (1 mL) and
NaCNBH3 (0.032 g, 0.5 mmol) were added, and the suspension
was stirred at rt for 20 h. Concd HCl (2 mL) was added, and
after being stirred for 15 min the reaction mixture was made
alkaline with excess solid K2CO3. Solids were removed by
filtration and washed several times with ether. The combined
filtrates were separated, and the aqueous layer was extracted
with ether. Chromatography (CH2Cl2/MeOH/concd NH4OH
85/15/1.5 f 75/25/2.5) gave first sparteine (9, 7.5 mg, 9%) and
second â-isosparteine (10, 9.0 mg, 11%). 10: 1H NMR (CDCl3)
δ 23.02 (dd, J ) 10.8, 6.6 Hz, 2H), 2.79 (ddd, J ) 12.6, 2.0, 2.0
Hz, 2H), 2.45 (ddd, J ) 12.6, 12.6, 2.5 Hz, 2H), 2.28 (bd, J )
11.8 Hz, 2H), 2.17 (dd, J ) 10.8, 2.7 Hz, 2H), 1.75 (m, 2H),
1.65 (m, 2H), 1.6 (m, 4H), 1.52 (dd, J ) 3.3, 3.3 Hz, 2H), 1.35
(m, 4H), 1.24 (m, 2H); 13C NMR (CDCl3) δ 62.8, 55.1, 54.9, 34.4,
28.6, 25.4, 22.6, 19.8; HRMS obsd mass 234.2106, calcd for
Red u ction of 26 to 28. Enamine 26 (0.277 g, 1 mmol) was
stirred with NaCNBH3 (0.094 g, 1.5 mmol) in a mixture of
acetonitrile (8 mL) and acetic acid (2 mL) for 20 h at rt. Concd
HCl (2 mL) was added, and after being stirred for 15 min the
reaction mixture was concentrated in vacuum, made alkaline
with K2CO3, and extracted with ether. Chromatography (CH2-
Cl2/MeOH/concd NH4OH 85/15/1.5) gave 28 (0.190 g, 68%) as
a syrup: 1H NMR (CDCl3) δ 7.72 (d, J ) 8.8 Hz, 1H), 7.71 (d,
J ) 8.9 Hz, 1H), 7.05 (d, J ) 7.3 Hz, 1H), 7.03 (d, J ) 7.3 Hz,
1H), 3.82, 3.81, 3.80, 3.78 (4 × CH3), 3.05-1.0 (m, 25H); IR
(CHCl3) 1585 cm-1
.
(()-Sp a r t ein e19,20 (9). Concd HCl (0.25 mL) was added
dropwise to a solution of oxime 26 (0.139 g, 0.5 mmol) in MeOH
(6 mL). The solution was cooled to -50 °C and treated with
ozone at 30 min intervals. The bath temperature was kept
between -45 and -55 °C, and after 7 h the solution was
purged with nitrogen followed by the addition of dimethyl
sulfide (0.19 mL, 2.5 mmol). After the solution was stirred at
C
15H26N2 234.2096.
J O9602130