Syntheses of benzoquinolizidine and benzoindolizidine derivatives as anti-amnesic agents

Article abstract of DOI:10.1016/S0968-0896(99)00078-4

Tacrine, one of the drugs available for Alzheimer's disease based on the cholinergic approach, suffers from considerable toxicity. Many analogues of tacrine have been prepared which retain the pharmacologically rich aminopyridine or aminoquinoline motifs. The current research was undertaken to produce an acetylcholinesterase inhibitor by employing 11-aminobenzoquinolizidines (4) and 10-aminobenzoindolizidines (5) as templates. Thus, we aimed to achieve three goals relative to tacrine: eliminate the pyridine and quinoline moieties and render the molecule less flat. Overall, the compounds we prepared were poorer inhibitors of acetylcholinesterase compared to tacrine. The single exception was compound 6f which exhibited an effect comparable to that of tacrine, but only at a dose of the order of 10^-3 M. However, despite the poor acetylcholinesterase inhibition by 6b, this compound proved to be an effective anti-amnesic agent at 45mg/kg dose. Copyright (C) 1999 Elsevier Science Ltd.

Full text of DOI:10.1016/S0968-0896(99)00078-4

Bioorganic & Medicinal Chemistry 7 (1999) 1637±1646
Syntheses of Benzoquinolizidine and Benzoindolizidine Derivatives
as Anti-amnesic Agents
Shikai Zhao, a Michael J. Totleben, a Jeremiah P. Freeman, a C. L. Bacon, b G. B. Fox, b
E. O'Driscoll, b A. G. Foley, b J. Kelly, b U. Farrell, b Ciaran Regan, b
Stephen A. Mizsak c and Jacob Szmuszkovicz a,*
aDepartment of Chemistry and Biochemistry, University of Notre Dame, Notre Dame, IN 46556, USA
bDepartment of Pharmacology, University College Dublin, Dublin, Ireland
cPharmacia and Upjohn, Kalamazoo, MI 49001, USA
Received 13 November 1998; accepted 25 February 1999
AbstractÐTacrine, one of the drugs available for Alzheimer's disease based on the cholinergic approach, su€ers from consider-
able toxicity. Many analogues of tacrine have been prepared which retain the pharmacologically rich aminopyridine or amino-
quinoline motifs. The current research was undertaken to produce an acetylcholinesterase inhibitor by employing 11-
aminobenzoquinolizidines (4) and 10-aminobenzoindolizidines (5) as templates. Thus, we aimed to achieve three goals relative to
tacrine: eliminate the pyridine and quinoline moieties and render the molecule less ¯at. Overall, the compounds we prepared were
poorer inhibitors of acetylcholinesterase compared to tacrine. The single exception was compound 6f which exhibited an e€ect
3
comparable to that of tacrine, but only at a dose of the order of 10 M. However, despite the poor acetylcholinesterase inhibi-
tion by 6b, this compound proved to be an e€ective anti-amnesic agent at 45 mg/kg dose. # 1999 Elsevier Science Ltd. All rights
reserved.
Introduction
The cholinergic approach to Alzheimer's disease
received an encouraging signal due to the recent
approval of tacrine hydrochloride (Cognex1) as the ®rst
drug for the treatment of this disease; it was launched in
1993. Tacrine (1) is a complex pharmacological agent1
which also inhibits the enzyme acetylcholinesterase
(AcChE), thus maintaining synaptic residence of acet-
ylcholine (AcCh).2 Two other AcChE inhibitors have
been marketed recently: donezepil (Aricept1, 2)3 and
rivastigmine (Exelon, 3).4 The de®ciency of tacrine as a
drug is related to liver toxicity and peripheral cholino-
mimetic actions.5 Tacrine belongs to the well known
structural class of aminopyridines6 which represent an
interesting group of potassium channel blockers, but are
toxic. It also incorporates the template of 4-aminoqui-
noline, well known for diverse pharmacological activ-
ity.7 Furthermore, tacrine possesses a relatively ¯at
aromatic structure which may be prone to intercalation.
Many analogues of tacrine have been prepared.8 Most
of these are structurally closely related to the parent
compound and retain the aminopyridine or aminoqui-
noline moiety.
Our approach to the problem of maintaining the
AcChE inhibitory activity while diminishing the toxicity
is based on the incorporation of the 11-amino benzo-
quinolizidine (4) and 10-aminobenzoindolizidine (5)
moieties as the templates for AcChE inhibitors. In 4
and 5 we have eliminated the pyridine and quinoline
* Corresponding author. Fax: +1-219-631-6652.
0968-0896/99/$ - see front matter # 1999 Elsevier Science Ltd. All rights reserved.
PII: S0968-0896(99)00078-4

1638
S. Zhao et al. / Bioorg. Med. Chem. 7 (1999) 1637±1646
hydrogenation gave the desired reduced product as a
single cisy isomer 6b (J=2.4 Hz, CH N). (The stereo-
selectivity in this reduction is consistent with that
reported for the reduction of 8a to alcohol 10).9
moieties while maintaining the two amino functionalities
and rendering the molecule less ¯at. We now report the
preparation of these two classes of compounds.
Chemistry
Both structures 6 and its oxygen analogue 7 were
derived from di€erent substituted benzoquinolizinones
8. Structure 6 could be obtained from 8 via an imine
The trans isomer 14 (J=8.6 Hz for CH N) was
obtained as the major product in the reduction of imine
9a with sodium metal in ethanol. The product ratio of
trans to cis is 3:2. As reported previously for the reduc-
tion of oximes,10 reaction with sodium in ethanol allows
thermodynamic control to give the more stable product
14 as the major isomer.
intermediate 9, and structure 7 could be obtained from
8 via alcohol 10.
The benzyl amine derivatives 6c and 6e were prepared
via the benzyl imine 9b and 9d followed by hydrogena-
tion under neutral conditions.
When 6c was subjected to further hydrogenation in
presence of hydrochloric acid, the benzyl group was
cleaved cleanly to give primary amine 6a.
The starting ketone 8a was prepared according to the
literature.9 N-Benzylation of ethyl pipecolinate hydro-
chloride (11) was followed by HCl-catalyzed hydrolysis of
the resulting ester 12a to give acid 13a which was cyclized
with polyphosphoric acid to give 1,3,4,11a-tetrahydro-
2H-benzo[b]quinolizin-11(6H)-one (8a). Ketones 8b and
8c are prepared similarly via ester 12b and 12c, and
acids 13b and 13c.
Attempts to make its trans-isomer 15 by sodium/ethanol
reduction of 9b failed; starting material was recovered.
However, 15 was obtained from the a-phenethyl imine
intermediate 16.11 Compound 16 was reduced to sec-
ondary amine 17 which was hydrogenated to give the
trans primary amine 15 (J=8.6 Hz for CH N).
y
We de®ne stereochemistry by the relationship of the bridgehead
hydrogen and the benzylic hydrogen on the adjacent carbon. In their
1H NMR spectra, the cis isomers show small J values. In the original
paper9 describing alcohol 10, the stereochemistry was de®ned by the
relationship of the bridgehead hydrogen and the hydroxyl group.
The amination of ketone 8a with methylamine was car-
ried out in methylene chloride via the imine inter-
mediate 9a; reduction by LiAlH4 (LAH) or catalytic

S. Zhao et al. / Bioorg. Med. Chem. 7 (1999) 1637±1646
1639
Ketone 8a was hydrogenated to alcohols 18 and 10, which
were converted to the corresponding acetates 20 and 7.
aminolactam 29 (J=9.46 Hz, CH NHCH3), again
containing ꢀ10% of the cis-isomer.
Chaki and co-workers found that the 4-acylaminopyridine
derivative 21 shows better choline uptake activity when
compared to tacrine.12 Structure±activity relationship stu-
dies led to compound 22 with even better activity.13
Attempts to prepare pure cis-isomer 30 by catalytic
hydrogenation of the methylimine of ketone 26 were
frustrated by lack of a good route to 26.14 The cis iso-
mer of 29, namely 29a, was obtained in low yield by
catalytic hydrogenation of imine 28.
Based on the similarity suggested above between tacrine
and 6a, we synthesized the butyramide 23 by the acylation
of 6a with butyryl chloride and triethylamine. In addition
compound 25 was prepared by the condensation of 2-oxo-
1-pyrrolidineacetic acid (24) and 6a in the presence of CDI.
Biological Studies
Overall, these benzoquinolizidine and benzoindolizidine
derivatives were poor inhibitors of acetylcholinesterase,
as compared to that observed with tacrine (Table 1).
The single exception was 6f, which exhibited an e€ect
comparable to that of tacrine, but only at a dose in the
order of 10 3M. In general, cholinesterase inhibition
was similar to that determined for acetylcholinesterase.
However, in some cases a stimulatory e€ect was
observed at lower concentrations (for example see com-
pounds 6a and 6b). Inhibition of acetylcholinesterase by
these benzoquinolizidine and benzoindolizidine deriva-
tives appeared to be competitive in nature as kinetic
studies with 6b, 5a and tacrine revealed an equivalent Ki
of 0.03 mM acetylthiocholine iodide substrate (Fig. 1).
However, the inhibitory concentration of tacrine (5 mM)
was 100-times lower than that observed for either 6b or
5a (500 mM) (data not shown for tacrine).
Because of the interesting activity of compound 6b, the
preparation of the analogous indolizidine 5a was
undertaken.
E€orts to cyclize N-benzylpyrrolidine-2-carboxylic acid,
analogous to that of N-benzylpipecolic acid, failed to
produce ketone 26. Therefore an indirect route to the
desired indolizidine derivatives 5 was developed.
Evaluation of the in vivo anti-amnesic action of these
derivatives con®rmed the lack of a tacrine-like anti-
acetylcholinesterase action as the majority of com-
pounds tested failed to reverse scopolamine-induced
amnesia of the passive avoidance response (Table 2).
However, despite the lack of acetylcholinesterase inhi-
bition by 6b, this compound proved to be an e€ective
anti-amnesic agent at concentrations of 45 mg/kg. Thus,
this benzoquinolizidine and benzoindolizidine series
would appear to include agents with an anti-amnesic
action that is independent of an anti-cholinesterase
e€ect. This suggestion is supported further by the lack
of in vivo e€ect with 6f despite this compound being an
order of magnitude more potent acetylcholinesterase
Benzindolizidine-3, 10-dione 2714 was converted to the
3-N-methylimine 28, which was then reduced with
lithium aluminum hydride to give 5a, (J=10.01 Hz,
CH N) contaminated with ꢀ10% of the cis-isomer 30.
Reduction of 28 with sodium borohydride produced

1640
S. Zhao et al. / Bioorg. Med. Chem. 7 (1999) 1637±1646
inhibitor than 6b. In general, all agents were devoid of any
overt behavioural toxicity, the single exception being 6a
which induced amnesia when administered alone.
carbon in CDCl3 solution. The predicted carbon spectra
were obtained from the ACD Labs CNMR program.
Peak positions are indicated in ppm down®eld from
internal TMS in d units. Mass spectra were obtained on a
MAT CH-5-DF (FAB), and Finnigan 8230 B (EI) mass
spectrometers. IR spectra were recorded on a Perkin±
Elmer 1420 Ratio Recording IR spectrophotometer.
Structure±Activity Relationship
The structure±activity relationship de®ning compound 6b
as the antiamnesic lead may be summarized as follows: (1)
the 6-6-6 ring system is preferred to 6-6-5; (2) cis stereo-
chemistry is preferred to trans; (3) the 11-amino group is
best as a secondary amine with a small substitutent such as
methyl; (4) unsubstituted aromatic ring is preferred.
Flash column chromatography was done on silica gel
(E. M. Merck silica gel 60, 230±400 mesh) in the stated
solvents. Melting points were obtained on a Thomas±
Hoover apparatus and are uncorrected. Product purities
were routinely checked by TLC. THF was tested for
peroxides (aqueous KI) prior to use and used without
further puri®cation or drying. All reactions were per-
formed under a nitrogen atmosphere in oven- or ¯ame-
dried glassware unless otherwise noted. The multi-
plicities of the 13C signals were determined by DEPT
experiments; s=C; d=CH; t=CH2; q=CH3. Methyla-
mine was obtained from either lecture bottle (Aldrich)
or aqueous solution (40%, Fluka) by distillation. Ethyl
pipecolinate was purchased from Aldrich.
Experimental
1H and 13C NMR spectra were recorded on a Varian
spectrometer at 300 MHz for proton and 75.4MHz for
Table 1. Inhibition of rat cholinesterase and acetylcholinesterase
activity by benzoquinolizidine and benzoindolizidine derivatives and
tacrine
Acetylcholinesterase
activitya
Cholinesterase
activitya
5
3
7
Test
Compounds
10 3 M 10
M
10 7 M 10
M
10 5 M 10
M
6a
6c
14
6b
18
10
20
7
6d
6e
6f
23
25
Tacrine
105.3
105.3
42.1
21.1
105.3
90.9
30.4
17.4
42.2
18.6
2.3
105.3
115.0
121.1
105.3
115.8
121.1
108.7
95.7
12.5
8.3
112.5
120.8
100.0
125.0
86.70
60.0
60.0
65.7
94.9
85.7
134.0
127.5
101.3
139.0
87.4
62.3
75.7
77.4
88.9
78.2
102.9
93.8
107.8
42.9
100.0
100.0
110.5
126.3
100.0
86.9
78.3
91.6
89.3
69.6
74.8
91.3
7.4
50.0
16.7
82.9
25.7
22.9
14.3
40.4
14.1
2.7
91.3
102.5
78.4
94.2
78.8
104.7
42.8
73.3
91.9
103.7
14.9
28.4
41.7
5.4
24.2
45.2
17.8
Figure 1. In¯uence of benzoquinolizidine and benzoindolizidine deri-
vatives on rat brain acetylcholinesterase activity. The data are illu-
strated as Lineweaver±Burke plots for control (closed squares) and 6b
(0.5 mM; closed circles) and 5a (0.5 mM; open circles).
a
Activity was determined as mmol/min/mg protein and is presented
as percent of the control value.
Table 2. Reversal of scopolamine-induced amnesia of a passive avoidance response by benzoquinolizidine and benzoindolizidine derivativesa
6a
(l0 mg/kg)
6a
(30 mg/kg)
6c
(10 mg/kg)
6b
(10 mg/kg)
6b
(30 mg/kg)
6b
(45 mg/kg)
6d
(10 mg/kg)
6f
(30 mg/kg)
Vehicle
600‹0
(6)
44‹10
(6)
148‹101
(5)
427‹121
(6)
44‹20
(5)
132‹103
(6)
357‹94
(6)
18‹5
(6)
48‹14
(14)
357‹94
(6)
21‹5
(6)
104‹48
(12)
368‹84
(6)
167‹86
(6)
258‹156
(5)
484‹67
(6)
37‹6
(6)
322‹112
(8)*
508‹100
(3)
41‹10
(6)
54‹20
(6)
409‹44
(9)
33‹6
(9)
142‹62
(9)
(0.9% saline)
Scopolamine
(0.8 mg/kg)
Test compound
+scopolamine
(0.8 mg/kg)
Test compound
alone
292‹92
(6)y
188‹102
(6)
410‹115
(6)
459‹53
(12)
529‹78
(6)
340‹146
(5)
592‹9
(6)
323‹65
(9)
a
Results are expressed as the mean‹SEM (n) of recall latency to enter the dark compartment of the passive avoidance apparatus. Signi®cant
reversal by the test analogues of the Scopolamine-induced recall de®cit is indicated by an asterisk (Pꢁ0.05). Recall de®cits induced by the test
compounds are indicated by y (Pꢁ0.05).

S. Zhao et al. / Bioorg. Med. Chem. 7 (1999) 1637±1646
1641
Ethyl N-(4-methoxyphenylmethyl)-pipecolinate (12b). p-
Methoxybenzyl chloride (55 mmol, 8.61 g) was added to
a mixture of K2CO3 (56 mmol, 7.74 g) and ethyl pipe-
colinate hydrochloride (50 mmol, 9.68 g) in DMF
(55 mL). The mixture was stirred for 3 days at 62ꢂC, and
was ®ltered. Ice-cold HCl solution (1.5 N, 50 mL) was
added to the ®ltrate and it was extracted with ether
(3Â100 mL). The aqueous phase was basi®ed with
NaOH solution (40%) and extracted with ether
(5Â100 mL). The combined extracts were washed with
brine, dried with Na2SO4 and MgSO4 and concentrated
21.95, 20.77; MS (FAB) 253 (2, M+), 208 (100), 125 (59);
.
Anal. calcd for C13H16ClNO2 HCl: C, 53.81; H, 5.90; N
4.83; Cl, 24.43; found C, 53.52; H, 6.14; N, 4.85; Cl, 24.15.
1,3,4,11a-Tetrahydro-2H-benzo[b]quinolizin-11(6H)-one
(8a). The following is a modi®cation of the procedure
reported.9 Acid 13a (5.10 g, 20.0 mmol) was placed in a
500-mL round-bottomed ¯ask. Polyphosphoric acid
(200 g) was added. The mixture was heated in an oil
bath gradually to 140 ꢂC with stirring. It was stirred at
140 ꢂC until the bubbling stopped, then cooled to rt and
poured into ice. The mixture was neutralized with aqu-
eous NaOH (40%) and extracted with ether (5Â80 mL).
The combined extracts were washed with brine
(2Â50 mL), dried (Na2SO4) and concentrated in vacuo
to give a red solid (3.08 g, 77%): 71±72 ꢂC (benzene, lit9
68±69 ꢂC); 1H NMR (300MHz) d 8.02 (dd, J=7.8, 1.3 Hz,
1H), 7.50 (td, J=7.3, 1.4 Hz, 1H), 7.35 (t, J=7.6Hz, 1H),
7.23 (d, J=7.6 Hz, 1H), 3.88 (d, J=15.2 Hz, H-6), 3.68 (d,
J=14.9 Hz, H-6), 3.09 (br d, J=11.2 Hz, 1H), 2.77 (br d,
J=10.5 Hz, 1H), 2.44 (m, 1H), 2.35 (td, J=11.3, 3.5 Hz,
1H), 1.90 (br d, J=12.5 Hz, 1 H), 1.32±1.76 (m, 4 H); 13C
NMR (75.4 MHz) d 195.87, 141.73, 133.53, 130.12, 127.34,
127.00, 126.17, 69.21, 57.14, 56.13, 26.66, 25.02, 23.81.
1
in vacuo to give 12b as a yellow oil (7.12 g): H NMR
(300MHz) d 7.23 (d, J=8.6 Hz, 2H), 6.84 (d, J=8.6Hz,
2H), 4.22 (q, J=7.1 Hz, 2H), 3.70 (s, Me), 3.73 (d,
J=13.1 Hz, 1H), 3.34 (d, J=13.6 Hz, 1H), 3.08 (dd,
J=8.2, 4.3 Hz, 1H), 2.93 (m, 1H), 2.10 (m, 2H), 1.80 (m,
2H), 1.56 (m, 3H), 1.30 (t, J=7.1Hz, 3H); 13C NMR d
174.00, 158.68, 130.43, 130.01, 113.46, 64.55, 60.30, 59.87,
55.20, 50.15, 29.64, 25.24, 22.67, 14.33; MS (EI) 277 (2,
M), 204 (90), 121 (100); HRMS (EI) m/e calcd for
C16H23NO3 277.1678, found 277.1682.
Ethyl N-(3-chlorophenylmethyl)pipecolinate (12c).
A
mixture of 3-chlorobenzyl chloride (11 mmol, 1.4 mL),
ethyl pipecolinate hydrochloride (10 mmol, 1.94 g) and
K2CO3 (12 mmol, 1.69 g) in DMF (10 mL) was stirred at
60 ꢂC for 3 days. It was then diluted with ether (50 mL)
and ®ltered. The ®ltrate was concentrated in vacuo to
1,3,4,11a-Tetrahydro-9-methoxy-2H-benzo[b]quinolizin-
11(6H)-one (8b). By using above procedure, N-(4-meth-
oxyphenylmethyl) pipecolinic acid hydrochloride (13b,
2.98 g) was converted to ketone 8b (950mg, 41%): mp
117±119 ꢂC (ether); 1H NMR (300MHz) d 7.48 (d,
J=2.7Hz, 1H), 7.14 (d, J=8.4 Hz, 1H), 7.08 (dd, J=8.4,
2.7 Hz, 1H), 3.84 (s, Me), 3.84 (d, J=14.8 Hz, 1H), 3.61
(d, J=14.8 Hz, 1H), 3.08 (br d, J=11.1 Hz, 1H), 2.74 (br
d, J=10.1 Hz, 1H), 2.42 (m 1 H), 2.34 (td, J=11.3, 3.6 Hz,
1H), 1.90 (m, 1H), 1.15±1.80 (m, 4H); 13C NMR
(75.4 MHz) d 195.89, 158.84, 134.60, 131.02, 127.51,
121.88, 108.96, 69.01, 56.61, 56.13, 55.52, 26.74, 25.03,
23.87; MS (El) 231 (64, M), 203 (53), 202 (53), 174 (10),
161 (83), 148 (27), 121 (100), 120 (31); Anal. calcd for
C14H17NO2: C, 72.70; H, 7.41; N, 6.06; found C, 72.47; H,
7.47; N, 5.91.
1
give 12c as a yellow oil (2.79g): H NMR (300MHz) d
7.21±7.35 (m, 4 H), 4.20 (q, J=7.1 Hz, 2 H), 3.77 (d,
J=13.6 Hz, 1H), 3.37 (d, J=13.6 Hz, 1 H), 3.14 (dd,
J=7.5Hz, 4.6 Hz, 1H), 2.92 (dt, J=11.9, 5.4 Hz, 1H),
2.15 (m, 1H), 1.34±1.90 (m, 6H), 1.29 (t, J=7.1 Hz, 3H);
13C NMR d 173.74 (s), 140.73 (s), 134.06 (s), 129.35 (d),
128.99 (d), 127.14 (d), 64.33 (d), 60.34 (t), 59.98 (t), 50.11
(t), 29.52 (t), 25.25 (t), 22.39 (t), 14.30 (q); MS (FAB) 282
(98, M+H), 208 (100); HRMS (FAB) m/e calcd for
(C15H20ClNO2+H) 282.1261, found 282.1248.
N-(4-Methoxyphenylmethyl)pipecolinic acid hydrochloride
(13b). A mixture of 12b (7.12 g) and concd HCl
(100 mL) was re¯uxed for 8 h. Solvent was removed in
vacuo and 2-propanol was added to the residue. Upon
cooling, a pale yellow solid was obtained as 13b (2.08 g):
mp 199±201 ꢂC (2-propanol); 1H NMR (300MHz, D2O) d
7.42 (d, J=8.1 Hz, 2 H), 7.05 (d, J=8.1Hz, 2 H), 4.45 (d,
J=13.1 Hz, 1 H), 4.09 (d, J=13.1 Hz, 1 H), 3.84 (s, Me);
MS (EI) 249 (2, M), 204 (38), 121 (100); Anal. calcd for
1,3,4,11a-Tetrahydro-8-chloro-2H-benzo[b]quinolizin-11
(6H)-one (8c). By using above procedure, N-(3-chlor-
ophenylmethyl) pipecolinic acid hydrochloride (13c,
901 mg) was converted to ketone 8c (451 mg, 64%)
which was crystallized from methanol to give a solid:
mp 106±108 ꢂC; 1H NMR (300MHz) d 7.94 (d,
J=8.4Hz, 1H), 7.32 (dd, J=8.4, 2.0 Hz, 1 H), 7.22 (m,
1H), 3.82 (d, J=15.3 Hz, 1H), 3.64 (d, J=15.3 Hz, 1H),
3.06 (br d, J=11.4 Hz, 1H), 2.76 (br d, J=10.5 Hz, 1H),
2.36 (m, 2H), 1.88 (m, 1H), 1.30±1.70 (m, 4H); 13C NMR
(75.4 MHz) d 194.85 (s), 143.70 (s), 139.77 (s), 128.68 (d),
128.65 (s), 127.82 (d), 126.15 (d), 68.97 (d), 56.69 (t), 56.01
(t), 26.58 (t), 24.97 (t), 23.70 (t); MS (El) 235 (51, M), 206
(63), 165 (100); Anal. calcd for C13H14ClNO: C, 66.24; H,
5.99; Cl, 15.04; N, 5.94; found C, 66.18; H, 6.02; Cl, 15.19;
N, 5.85.
.
.
C14H19NO3 HCl 0.2H2O: C, 58.11; H, 7.11; N 4.84; Cl,
12.25; found C, 58.02; H, 7.21; N, 4.64; Cl, 12.57. The
mother liquor was concentrated to give a foam (5.59 g)
which was identical to 13b by 1H NMR.
N-(3-Chlorophenylmethyl)pipecolinic acid hydrochloride
(13c). Using the same procedure as above, ethyl N-(3-
chlorophenylmethyl)-pipecolinate (12c, 2.79g) was con-
verted to 13c (2.78 g): 221±223 ꢂC (MeOH:2-propanol);
1H NMR (300MHz, D2O) d 7.40±7.55 (m, 4 H), 4.54 (d,
J=12.9 Hz, 1 H), 4.13 (d, J=13.2 Hz, 1 H), 3.88 (dd,
J=11.9 Hz, 3.5 Hz, 1 H), 3.48 (br d, J=12.9 Hz, 1 H), 3.02
(td, J=12.6, 3.0 Hz, 1 H), 2.31 (m, 1 H), 1.50±1.90 (m, 5
H); 13C NMR (75.4 MHz, D2O) d 172.08, 134.25, 131.24,
130.67, 130.31, 130.25, 129.87, 65.37, 59.12, 51.66, 27.80,
Methylimine 9a. A solution of ketone 8a (1.10 g,
5.47 mmol) in CH2Cl2 (20 mL) was added to methyla-
mine (5 mL) followed by the addition of TiCl4
(2.74 mmol, 2.74 mL of 1.0 M solution in toluene). The

1642
S. Zhao et al. / Bioorg. Med. Chem. 7 (1999) 1637±1646
mixture was stirred at rt for 10 h. It was ®ltered through
Celite and rinsed with CH2Cl2. The ®ltrate was con-
centrated in vacuo to give a mixture of red solid and oil
(1.29 g, 100%) which was directly used for the next step
without puri®cation.
The amine was converted to the hydrochloride and crys-
tallized from MeOH:2-PrOH:ether to give a gray solid:
mp >230 C; Anal. calcd for C14H19ClN2 2HCl 0.2C3-
H8O: C, 52.24; H, 6.79; Cl, 31.68; N, 8.34; found C, 52.04;
H, 6.51; Cl, 31.36; N, 8.25.
ꢂ
.
.
cis-N-Methyl-1,3,4,6,11,11a-hexahydro-11-amino-2H-
benzo[b]quinolizidine (6b). Imine 9a (3.05 mmol, based on
starting ketone 8a was dissolved in THF (20 mL). LAH
(380mg) was added gradually to the mixture. The sus-
pension was heated at re¯ux for 0.5 h and then stirred at rt
for 10h and quenched by the successive addition of H2O
(380mL), NaOH (aqueous, 15%, 380 mL) and H2O
(1.1 mL). The mixture was ®ltered and rinsed with ether.
The ®ltrate was concentrated to give 6b as a red oil
(513mg, 79% based on starting ketone 8a), identical by
1H NMR to the compound obtained below by catalytic
hydrogenation.
trans-N-Methyl-1,3,4,6,11,11a-hexahydro-2H-benzo[b]
quinolizin-11-amine (14). The imine 9a (1.29 g, 5.47mmol
based on ketone 8a) was dissolved in absolute ethanol
(20 mL). Sodium (2.23 g) was added in small pieces to
maintain the reaction at re¯ux. After all the sodium was
added, more ethanol (10 mL) was added to complete the
reaction. The mixture was then cooled to 0 ꢂC and water
(10 mL) was slowly added. Ethanol was removed on a
rotary evaporator. The residue was extracted with ether
(4Â40mL), and the combined extracts were washed with
brine (2Â20mL), dried (Na2SO4) and concentrated in
1
vacuo to give a red oil (780mg, 66%). H NMR showed
that the product contained about 30% starting material. It
was reduced again by the addition of sodium (1.43 g) to
the solution of the oil in ethanol (20 mL). The reaction
mixture was quenched by the addition of water (10 mL).
Aqueous HCl solution (3 N, 50mL) was added and the
mixture was extracted with ether (50 mL). The aqueous
layer was basi®ed by the addition of solid NaOH, and was
extracted with ether (3Â50mL). The combined extracts
were dried and concentrated in vacuo to give a red oil
(659mg, 56%, trans:cis=3:2 by NMR). The oil was
separated on radial chromatography eluting with
CH2Cl2:MeOH:NH4OH (150:8:1) to give two fractions
(339mg, trans:cis=3:2; 30mg, trans:cis=9:1). The NMR
below was run on the latter fraction.
Catalytic hydrogenation of imine 6a (2.39 mmol, based
on the starting ketone 8a) in absolute ethanol (10 mL)
with Pd on carbon (10%, 210 mg) for 12 h at a pressure
of 46 psi gave 6b as an oil (421mg, 82% based on starting
ketone 8a): 1H NMR (300MHz) d 7.05±7.22 (m, 4 H), 4.00
(d, J=16.0 Hz, H-6), 3.34 (d, J=16.0 Hz, H-6), 3.24 (d,
J=2.4Hz, 1H), 3.07 (br d, J=11.3 Hz, 1H), 2.39 (dt,
J=11.2, 2.8 Hz, 1H), 2.36 (s, CH3), 2.10 (m, 2H), 1.94 (m,
1H, NH), 1.83 (m, 1H), 1.70 (m, 3H), 1.33 (m, 1H); 13C
NMR (75.4 MHz) d 136.26, 134.10, 128.88, 126.83, 126.23,
125.13, 61.80, 60.43, 58.32, 56.67, 34.36, 28.20, 25.57,
24.37; HCl salt mp: >230ꢂC (MeOH:2-PrOH:ether); MS
(FAB), m/e 217 (100, M+H), 186 (36); HRMS (FAB) m/e
calcd for C14H20N2+ H) 217.1705, found 217.1704; Anal.
1
.
calcd for C14H20N2 2HCl: C, 58.14; H, 7.67; Cl, 24.51; N,
9.68. Found: C, 58.21; H, 7.86; Cl, 24.35; N, 9.58.
14. H NMR (300 MHz) d 7.39 (d, J=7.6 Hz, 1H), 7.22
(t, J=7.3 Hz, 1H), 7.14 (td, J=7.4, 1.0 Hz, 1H), 7.03 (d,
J=7.3 Hz, 1H), 3.72 (d, J=14.9 Hz, 1H), 3.64 (d,
J=8.4 Hz, 1H), 3.39 (d, J=15.1 Hz, 1H), 3.05 (br d,
J=11.3 Hz, 1H), 2.30 (s, CH3), 1.25±2.30 (m, 9H); 13C
NMR (75.4 MHz) d 136.08, 135.48, 126.94, 126.61,
126.64, 125.84, 62.55, 61.86, 58.01, 56.32, 31.81, 31.05,
25.25, 24.23; MS (FAB), m/e 217 (100, M+H), 186 (29);
HRMS (FAB) m/e calcd for (C14H20N2O+H) 217.1705,
found 217.1708.
In the same way 6d was prepared from ketone 8b via imine
9b and was obtained as an oil: 1H NMR (300MHz) d 6.97
(d, J=8.4 Hz, 1H), 6.77 (dd, J=8.4 2.7 Hz, 1H), 6.67 (d,
J=2.6Hz, 1H), 3.93 (d, J=15.4 Hz, 1H), 3.79 (s, OCH3),
3.26 (d, J=15.3 Hz, 1H), 3.18 (d, J=2.4 Hz, 1H), 3.05 (br
d, J=11.3 Hz, 1H), 2.38 (s, NCH3), 1.25±2.36 (m, 8 H);
13C NMR d 157.07, 137.61, 127.15, 126.18, 113.59, 112.89,
61.83, 60.75, 57.79, 56.68, 55.19, 34.57, 28.13, 25.56, 24.37;
MS (FAB) 247 (100, M+H), 216 (88), 163 (66); HCl salt
mp: >230ꢂC (MeOH:2-PrOH:ether); Anal. calcd for
cis-N-Benzyl-1,3,4,6,11,11a-hexahydro-11-amino-2H-
benzo-[b]quinolizidine (6c). Benzylamine (1.2 mL,
10.76 mmol) was added to a solution of ketone 8a
(721 mg, 3.59 mmol) in CH2Cl2 (20 mL) followed by the
addition of TiCl4 (1.8 mmol, 1.8 mL of 1.0 M solution in
toluene). The mixture was stirred at rt for 20 h. It was
®ltered through Celite and the solid rinsed with CH2Cl2.
The ®ltrate was concentrated in vacuo to give 9b as a
yellow solid which was directly used for the next step
without puri®cation.
.
.
C15H22N2O 2HCl 0.2H2O: C, 55.80; H, 7.62; Cl, 21.96; N,
8.68; found C, 56.15; H, 7.59; Cl, 21.64; N, 8.56.
6f. The imine 9e (1.06 g) was dissolved in THF (10 mL),
Pt/C sul®ded (Aldrich) (410 mg) was added and the
mixture was hydrogenated (50 psi) for 18 h at rt. It was
®ltered and concentrated. The residue was chroma-
tographed on silica gel eluting with CH2Cl2:MeOH:N-
H4OH (150:8:1) to give 6f as a red oil (287mg, 27%): 1H
NMR (300MHz) d 7.07 (m, 3H), 3.93 (d, J=16.2 Hz,
1H), 3.28 (d, J=16.5 Hz, 1H), 3.22 (d, J=2.4Hz, 1H),
3.04 (br d, J=10.8 Hz, 1H), 2.32 (s, CH3), 1.20±2.10 (m,
8H); 13C NMR d 135.93 (s), 134.54 (s), 132.35 (s), 130.13
(d), 126.03 (d), 125.29 (d), 61.51 (d), 59.64 (d), 57.79 (t),
56.43 (t), 34.01 (q), 27.93 (t), 25.39 (t), 24.17 (t); MS (FAB)
251 (78, M+H), 154 (100), 136 (77); HRMS (FAB) m/e
calcd for (C14H19ClN2+H) 251.1315, found 251.1319;
The benzylimine 9b (189 mg) was dissolved in absolute
ethanol (5 mL) and palladium on carbon (10%, 74 mg)
was added. The mixture was hydrogenated in a Parr for
5 h at a pressure of 46 psi and then ®ltered through
Celite. The ®ltrate was concentrated to give 6c as a yel-
1
low oil (93 mg, 50%): H NMR (300 MHz) d 7.01±7.40
(m, 9 H), 4.00 (d, J=15.9 Hz, 1H), 3.86 (d, J=13.8 Hz,
1H), 3.69 (d, J=13.5 Hz, 1H), 3.336 (d, J=15.9 Hz,

S. Zhao et al. / Bioorg. Med. Chem. 7 (1999) 1637±1646
1643
1H), 3.335 (d, J=1.8 Hz, 1H), 3.08 (br d, J=11.4 Hz,
1H), 1.25±2.42 (m, 8H); 13C NMR (75.4 MHz) d 141.17,
137.05, 134.17, 128.57, 128.24, 128.12, 126.66, 126.55,
126.31, 125.22, 62.15, 58.42, 57.36, 56.76, 50.92, 28.15,
25.65, 24.44; MS (FAB), m/e 293 (100, M+H), 209 (6),
198 (36), 186 (61), 91 (30); HRMS (FAB) m/e calcd for
(C2H2N2+H) 293.2018, found 293.2009. HCl salt was
prepared from ethereal HCl solution and crystallized
from MeOH:2-PrOH:ether: mp 185±188 ꢂC; Anal. calcd
(4 mL) at 0 ꢂC. After 30 min., it was warmed to rt and
stirred overnight. The reaction was quenched by the
addition of water and MeOH was removed in vacuo.
The residue was extracted with ether, dried (Na2SO4)
and concentrated in vacuo to give an oil. The oil was
dissolved in EtOH (5 mL) and HCl (3 N, 2 mL) and
hydrogenated for 12 h with a Parr apparatus (50 psi)
catalyzed by Pd/C (201 mg). It was worked up and the
crude product was puri®ed by chromatography (silica
gel column eluting with CH2Cl2:MeOH:NH4OH) to give
.
for C20H24N2 HCl: C, 73.04; H, 7.66; Cl, 10.78; N, 8.52.
Found: C, 72.73; H, 7.72; Cl, 11.11; N, 8.14.
1
15 as a red oil (47 mg): H NMR (300MHz) d 7.51 (d,
J=7.7Hz, 1H), 7.22 (d, J=7.4 Hz, 1H), 7.15 (d,
J=7.3Hz, 1H), 7.01 (d, J=7.5 Hz, 1H), 3.78 (d, J=15.1
Hz, 1H), 3.67 (d, J=8.6 Hz, 1H), 3.43 (d, J=15.1 Hz, 1H),
3.05 (br d, J=11.2 Hz, 1H), 2.30 (m, 1H), 2.16 (dt,
J=11.4, 3.9 Hz, 1H), 1.24±1.96 (m, 8H); 13C NMR
(75.4 MHz) dd 138.54, 134.00, 126.84, 126.72, 126.41,
125.61, 67.60, 58.29, 56.20, 55.55, 30.89, 25.38, 24.11; MS
(FAB), m/e 203 (100, M+H), 186 (45); HRMS (FAB) m/e
calcd for (C13H18N2+H) 203.1548, found 203.1531.
6e. Compound 6e was obtained as a brown oil (313 mg,
1
50%) from corresponding imine 9d (614 mg): H NMR
(300 MHz) d 7.20±7.40 (m, 5H), 6.97 (d, J=8.7 Hz, 1H),
6.76 (dd, J=8.4, 2.7 Hz, 1H), 6.52 (d, J=2.7 Hz, 1H),
3.97 (d, J=15.6 Hz, 1H), 3.88 (d, J=13.5 Hz, 1H), 3.76
(s, CH3O), 3.74 (d, J=13.5 Hz, 1H), 3.32 (d, J=1.8 Hz,
1H), 3.28 (d, J=15.6 Hz, 1H), 3.09 (br d, J=11.1 Hz,
1H), 1.30±2.40 (m, 8 H); 13C NMR (75.4 MHz) d 157.27,
141.13, 138.07, 128.33, 128.21, 127.26, 126.65, 113.16,
62.23, 57.83, 57.46, 56.79, 55.25, 51.18, 28.12, 25.60, 24.41,
two aromatic signals are not discernible due to overlap
(one quaternary, one CH); MS (FAB), m/e 323 (100,
M+H), 239 (4), 216 (39), 108 (3), 91 (10). HCl salt was
prepared from ethereal HCl solution and crystallized from
MeOH:2-PrOH: mp 213±215 ꢂC; Anal. calcd for
trans-1,3,4,6,11,11a-Hexahydro-2H-benzo[b]quinolizin-
11-ol (18) and cis-1,3,4,6,11,11a-hexahydro-2H-benzo[b]
quinolizin-11-ol (10). A mixture of ketone 8a (0.97 g,
4.83 mmol) and Pd/C (10%, 1.50 g) in THF (50 mL) and
HCl (0.5 N, 5 mL) was hydrogenated in a Parr appara-
tus (45 psi) for 15 h. The mixture was basi®ed with
NaOH (2 N) and extracted with ether. The extract was
dried (Na2SO4) and concentrated in vacuo to give a
solid (790 mg). The solid was recrystallized from ben-
zene to give 18 (410 mg) as a colorless solid, mp 158±
160 ꢂC (lit.9 162 ꢂC). The mother liquor was evaporated
and the residue was crystallized from ethanol to give 10
(132 mg) as a colorless solid, mp 152±154 ꢂC.
.
C21H26N2O 2HCl: C, 63.80; H, 7.14; Cl, 17.93; N, 7.09.
Found: C, 63.83; H, 6.92; Cl, 17.47; N, 7.02.
cis-1,3,4,6,11,11a-Hexahydro-11-amino-2H-benzo[b]qui-
nolizidine (6a). Pd on carbon (10%, 154 mg) was added
to a solution of N-benzylamine 6c (93 mg) in THF
(10 mL) and aqueous HCl (0.5 N, 1 mL). The mixture
was hydrogenated in a Parr. Work up, as above, gave 6a
1
1
as a colorless oil (29 mg): H NMR (300 MHz) d 7.00±
18. H NMR (300 MHz) d 7.52 (d, J=7.3 Hz, 1H), 7.22
7.25 (m, 4H), 3.93 (d, J=15.7 Hz, 1H), 3.58 (d,
J=2.6 Hz, 1H), 3.33 (d, J=15.7 Hz, 1H), 3.01 (br d,
J=11.3 Hz, 1H), 2.32 (dt, J=10.8, 3.0 Hz, 1H), 2.12 (td,
J=11.6, 3.6 Hz, 1H), 1.25±1.93 (m, 8H); 13C NMR
(75.4 MHz) d 139.86, 133.60, 128.66, 126.73, 126.46,
125.85, 61.64, 58.60, 56.57, 53.11, 28.40, 25.70, 24.37; MS
(FAB), m/e 203 (100, M+H), 186 (33), 84 (10); HRMS
(FAB) m/e calcd for (C13H18N2+H) 203.1548, found
203.1560; the hydrochloride was prepared with ethereal
HCl and was crystallized from MeOH:2-PrOH:ether: mp
(m, 2 H), 7.03 (d, J=7.3Hz, 1H), 4.47 (d, J=8.4 Hz, 1H),
3.81 (d, J=15.3 Hz, 1H), 3.44 (d, J=15.3 Hz, 1H), 3.04
(br d, J=11.3 Hz, 1H), 2.33 (m, 1H), 2.18 (td, J=11.5,
3.9 Hz, 1H), 2.07 (m, 1H), 1.30±1.90 (m, 6H).
1
10. H NMR (300 MHz) d 7.35 (m, 1H), 7.22 (m, 2H),
6.94 (m, 1H), 4.12 (s, 1H), 3.56 (s, OH), 3.39 (d,
J=15.7 Hz, 1H), 3.16 (d, J=15.6 Hz, 1H), 2.93 (br d,
J=10.9 Hz, 1H), 2.24 (dt, J=11.2, 2.7 Hz, 1 H), 1.20±
2.10 (m, 7 H); 13C NMR (75.4 MHz) d 137.11, 134.04,
129.41, 127.54, 126.52, 126.08, 69.67, 62.39, 58.02,
56.14, 27.36, 25.23, 24.07; MS (FAB), m/e 204 (100,
M+H), 186 (40), 84 (50); Anal. calcd for C13H17NO: C,
76.81; H, 8.43; N, 6.89. Found: C, 76.94; H, 8.49; N,
6.74.
ꢂ
.
.
.
>230 C; Anal. calcd for C13H18N2 2HCl 0.6H2O 0.2
isopropanol: C, 54.84; H, 7.65; Cl, 23.81; N, 9.41. Found:
C, 54.86; H, 7.69; Cl, 23.80; N, 9.32.
trans-1,3,4,6,11,11a-Hexahydro-11-amino-2H-benzo[b]
quinolizidine (15). (‹)-a-Methylbenzylamine (0.99 mL,
7.70 mmol) was added to a solution of ketone 8a
(516 mg, 2.57 mmol) in CH2Cl2 (20 mL) followed by the
addition of TiCl4 (1.28 mmol, 1.28 mL of 1.0 M solution
in toluene). The mixture was stirred at rt for 20 h. Hex-
anes (20 mL) was added and the suspension was ®ltered
through Celite. The ®ltrate was concentrated in vacuo
to give a red oil (719 mg) which was directly used for the
next step without puri®cation.
trans-1,3,4,6,11,11a-Hexahydro-11-acetyl-2H-benzo[b]
quinolizidine (20).
A solution of AcCl (238 mL,
3.34 mmol) in CH2Cl2 (2 mL) was added to a solution of
alcohol 18 (563 mg, 2.78 mmol) and Et3N (465 mL,
3.34 mmol) in CH2Cl2 (10 mL) at 0 ꢂC. The mixture was
stirred at rt for 2 h and then diluted with CH2Cl2
(50 mL). The solution was washed with Na2CO3 (satd,
10 mL) and H2O (10 mL), dried (Na2SO4) and con-
centrated in vacuo. The residue was chromatographed
on silica gel column eluting with CHCl3:MeOH:N-
H4OH (99:0.8:0.2) followed by crystallization (EtOAc:
Part of above oil (464 mg) was dissolved in MeOH
(1 mL) and added to a NaBH4 suspension in MeOH

1644
S. Zhao et al. / Bioorg. Med. Chem. 7 (1999) 1637±1646
hexanes) to give 20 as a yellow solid: mp 116±117 ꢂC;
1H NMR (300 MHz) d 7.03±7.22 (m, 4H), 5.97 (d,
J=8.6 Hz, H-11), 3.82 (d, J=15.3 Hz, H-6), 3.49 (d,
J=15.1, H-6), 3.06 (br d, J=11.2 Hz, 1H), 2.20±2.35
(m, 2H), 2.17 (s, CH3), 1.64±1.96 (m, 3H), 1.25±1.47 (m,
3H); 13C NMR (75.4 MHz) d 171.16, 135.06, 132.43,
129.56, 128.35, 126.73, 125.80, 70.00, 61.21, 58.09,
56.48, 27.40, 25.36, 24.07, 21.24; MS (EI), m/e 246 (87,
stirred at rt for 24 h. The mixture was acidi®ed by the
addition of HCl (1 N) and extracted with CHCl3
(3Â20 mL). The combined extracts were dried (Na2SO4)
and concentrated in vacuo to give acid 24 as a colorless
solid: mp 140±142 ꢂC; H NMR (300 MHz) d 8.92 (br s,
1
OH), 4.09 (s, 2 H), 3.53 (t, J=7.1 Hz, 2H), 2.48 (t,
J=8.1 Hz, 2H), 2.10 (quin, J=7.5 Hz, 2H); 13C NMR
(75.4 MHz) d 176.72, 171.59, 48.13, 44.19, 30.35, 17.90;
MS (FAB), m/e 144 (100, M+H), 98 (19); Anal. calcd
for C6H9NO3: C, 50.35; H, 6.34; N, 9.79. Found: C,
49.99; H, 6.21; N, 9.78.
.
M+H), 186 (100); Anal. calcd for C15H19NO2 0.1H2O:
C, 72.91; H, 7.83; N, 5.67. Found: C, 72.87; H, 7.73; N,
5.60.
cis-1,3,4,6,11,11a-Hexahydro-11-acetoxy-2H-benzo[b]
quinolizidine (7). A solution of AcCl (82 mg, 1.05 mmol)
in CH2Cl2 (1 mL) was added to a solution of alcohol 10
(193 mg, 0.95 mmol) and Et3N (146 mL, 1.05 mmol) in
CH2Cl2 (5 mL) at rt. The mixture was stirred at rt for
3 h and then diluted with CH2Cl2 (50 mL). The solution
was washed with Na2CO3 (satd, 10 mL) and H2O
(10 mL), dried (Na2SO4) and concentrated in vacuo.
The residue was chromatographed on silica gel column
eluting with CHCl3:MeOH:NH4OH (99:0.8:0.2) fol-
lowed by crystallization (EtOAc:hexanes) to give 7 as a
yellow solid (59 mg, 26%): mp 92±94 ꢂC; 1H NMR
(300 MHz) d 7.16±7.34 (m, 3H), 7.07 (d, J=7.4 Hz, 1H),
5.98 (d, J=2.8 Hz, H-11), 4.00 (d, J=15.6 Hz, H-6),
3.35 (d, J=15.6, H-6), 3.19 (br d, J=11.5 Hz, 1H), 2.45
(dt, J=10.8, 3.1 Hz, 1H), 2.11 (s, CH3), 1.85 (m, 1 H),
1.50±1.75 (m, 5 H), 1.35 (m, 1 H); 13C NMR (75.4 MHz)
d 171.16, 135.06, 132.43, 129.56, 128.35, 126.73, 125.80,
70.00, 61.21, 58.09, 56.48, 27.40, 25.36, 24.07, 21.24; MS
(EI), m/e 244 (6, M H), 202 (54), 185 (100), 156 (23),
143 (13), 129 (15), 120 (60); HRMS (EI) m/e calcd for
(C15H19NO2 H) 244.1338, found 244.1330; Anal. calcd
A mixture of 24 (2.52 mmol, 360 mg) and carbonyldi-
imidazole (2.52 mmol, 408 mg) in THF (20 mL) was
stirred at rt for 2 h. Compound 6a in THF (5 mL) was
added in 5 min. The mixture was stirred at rt for 40 h. It
was acidi®ed by the addition of HCl (0.5 N, 20 mL) and
washed with ether (2Â30 mL). The aqueous layer was
basi®ed by the addition of solid NaOH, and extracted
with ether (3Â50 mL). The combined extracts were
dried (Na2SO4) and concentrated to give a brown solid.
The solid was washed with ethyl acetate to give 25 as an
o€-white solid (387 mg, 47%): mp 174±176 ꢂC (EtOAc);
1H NMR (300 MHz) d 7.32 (br d, J=7.4 Hz, 1H), 7.13±
7.23 (m, 2H), 7.02 (br d, J=6.9 Hz, 1H), 6.46 (d,
J=9.6 Hz, NH), 4.97 (dd, J=9.6, 2.4 Hz, 1H), 3.83±4.00
(m, 3H), 3.36±3.54 (m, 2H), 3.36 (d, J=15.6 Hz, 1H),
3.08 (br d, J=11.1 Hz, 1H), 1.26±2.49 (m, 12H); 13C
NMR (75.4 MHz) d 175.56, 166.91, 135.52, 133.94,
128.97, 127.55, 126.69, 125.79, 60.34, 58.35, 56.15,
50.05, 48.01, 46.72, 30.33, 27.87, 25.55, 23.80, 17.97; MS
(FAB), m/e 328 (100, M+H), 185 (38); Anal. calcd for
C19H25N3O2: C, 69.70; H, 7.70; N, 12.83. Found: C,
69.46; H, 7.73; N, 12.69.
.
for C15H19NO2 H2O: C, 72.91; H, 7.83; N, 5.67. Found:
C, 72.81; H, 7.56; N, 5.56.
trans-10-N-Methylaminobenzindolizidine (5a). Methyla-
mine (lecture bottle, Fluka) was condensed at 78ꢂC into
a ¯ask containing the ketone 271 (8.8mmol, 1.756 g) and
CH2Cl2 (30 mL) was added. TiCl4 (4.4 mmol, 4.4 mL of 1
M solution in toluene) was gradually added at 78ꢂC and
the mixture was stirred at 78ꢂC for 30min, then at rt for
22h under N2. Petroleum ether (bp 35±60 ꢂC) (80 mL) was
added and then the mixture was ®ltered through Celite.
The ®ltrate was concentrated in vacuo to give 28 as a red
N-Butyryl 6a (23). Butyryl chloride (2.82 mmol,
295 mL) in CH2Cl2 (5 mL) was added dropwise to a
solution of 6a (2.56 mmol, 518 mg) and triethylamine
(2.82 mmol, 393 mL) in CH2Cl2 (40 mL). The mixture
was stirred overnight at rt. Saturated Na2CO3 solution
(10 mL) was added and the aqueous layer was extracted
with additional CH2Cl2 (2Â30 mL). The combined
extracts were dried (Na2SO4) and concentrated. The
residue was chromatographed on silica gel eluting with
CH2Cl2:MeOH:NH4OH (150:8:1) to give 23 as a brown
solid (299 mg, 43%): mp 141±143 ꢂC; 1H NMR
(300 MHz) d 7.34 (dd, J=6.6, 2.4 Hz, 1H), 7.14±7.23
(m, 2H), 7.02 (dd, J=6.6, 2.1 Hz, 1H), 6.10 (d,
J=9.6 Hz, 1H), 5.06 (dd, J=9.9, 2.7 Hz, 1H), 3.92 (d,
J=15.6 Hz, 1H), 3.37 (d, J=15.6 Hz, 1H), 3.09 (br d,
J=11.4 Hz, 1H), 2.44 (dt, J=10.8, 3.0 Hz, 1H), 2.40 (m,
3H), 1.25±1.80 (m, 8H), 0.91 (t, J=7.4 Hz, 3H); 13C
NMR (75.4 MHz) d 171.98, 135.98, 133.78, 129.19,
127.36, 126.75, 125.67, 60.78, 58.46, 56.37, 49.48, 38.85,
27.89, 25.57, 23.86, 19.21, 13.75; MS (FAB), m/e (100,
M+H); Anal. calcd for C17H24N2O: C, 74.96; H, 8.88;
N, 10.28. Found: C, 74.85; H, 8.67; N, 10.01.
1
viscous oil (1.63 g). H NMR (CDCl3): d 1.90 (m, 1H),
2.42 (m, 1H), 2.56 (m, 1H), 2.79 (m, 1H), 3.45 (s, 3H), 4.04
(d, 1H), 4.87 (m, 1H), 5.14 (d, 1H), 7.15±7.44 (m, 3H), 8.00
(d, 1H).
Crude imine 28 (880 mg, 4.1 mmol) in 12 mL of THF
was added dropwise to a suspension of LiAlH4 (623 mg,
4.1 mmol) in 12 mL of THF at rt under N2. After gently
re¯uxing for 15.5 h, the mixture was cooled to rt, quen-
ched by successive addition of water (0.88 mL), 15%
NaOH (0.88 mL), and water (2.6 mL). This mixture was
®ltered through Celite and concentrated. The residue
was dissolved in CHCl3, dried over Na2SO4, ®ltered and
concentrated to give 600 mg of a dark red viscous resi-
due. It was dissolved in CHCl3 and ¯ash chromato-
graphed on silica in CH2Cl2:MeOH:NH4OH (95:4:1) to
1
N-[ꢀ-N-(2-oxopyrrolidinyl)acetyl 6a (25). A solution of
methyl 2-oxo-1-pyrrolidineacetate (10 mmol, 1.57 g,
Aldrich) in THF (20 mL) and NaOH (1 N, 10 mL) was
give 547 mg of a dark red viscous residue, 5a: H NMR
(CDCl3) d 1.69 (m, 1H), 1.87 (m, 2H), 2.28 (m, 1H),
2.41 (s, 3H), 2.47 (m, 1H), 2.58 (m, 1H), 3.17 (m, 1H),

S. Zhao et al. / Bioorg. Med. Chem. 7 (1999) 1637±1646
1645
3.62 (d, J=14.0 Hz, 1H), 3.79 (d, J=10.01Hz, 1H), 4.03
(d, J=14.03Hz, 1H), 7.09 (d, J=7.63 Hz, 1H), 7.17 (t,
J=7.76, 15.13 Hz, 1H), 7.25 (t, J=7.23, 14.47 Hz, 1H),
7.52 (d, J=7.63 Hz, 1H), NH is not discernible and prob-
ably appears between d 1 and d 4; 13C NMR (CDCl3) d
21.57, 30.04, 32.33, 54.71, 55.85, 62.70, 64.92, 126.20 (2C),
126.37, 126.63, 135.83, 137.24; IR (neat ®lm) 3285
(broad), 3060, 3020, 2950, 2870, 2795, 1735±1565 (broad)
1475, 1450, 1155, 1035, 920, 905, 780, 720 cm 1; MS
(FAB) m/e 203 (M+H, 74.8%), 201 (74.1), 172
(M NHCH3, 100), 170 (85), 133 (53.1), 132 (27), 118
(17.7). HRMS (FAB): C13H18N2, calcd 202.1469 Found
203.1555 M+H).
modi®cations described by Whittaker.17 Brie¯y, the
assay was carried out at room temperature using 0.1 M
phosphate bu€er, pH 8.0, containing 0.075 M acet-
ylthiocholine iodide, 0.01 M 5,5-dithio-bis-2-nitro-
benzoic acid (DTNB) and, in the case of
acetylcholinesterase determinations, 0.03 M ethopropa-
zine. The DTNB was prepared in 0.1 M phosphate
bu€er, pH 7.0, which contained 0.018 M sodium bicar-
bonate to ensure stability. The ethopropazine was pre-
pared in ethanol, the concentration of which never
exceeded 0.3%. The reaction was started by the addition
of 15 mL of enzyme preparation to a ®nal assay volume
of 1.585 mL, mixed thoroughly and progress was mon-
itored continuously for 5 min at 412 nM in a Philips PU
8625 UV/VIS spectrophotometer. All solutions were
freshly prepared and protected from light. The acetyl-
thiocholine, DTNB, and ethopropazine were obtained
from Sigma Chemical Co. Ltd. (UK). All other
reagents were of the highest grade available from rou-
tine suppliers.
trans-10-N-Methylaminobenzindolizidin-3-one (29). Crude
imine 28 (800 mg, 3.7 mmol) was dissolved in 13 mL of
absolute ethanol and cooled to 0 ꢂC. The cold solution
was treated with portions of NaBH4 (560 mg,
14.8 mmol) with stirring at 0 ꢂC for 30 min, then at rt for
18 h. Quenching with water (20 mL) followed by extrac-
tion with CH2Cl2 (4Â20 mL), drying and concentration
gave 677 mg of residue, which was chromatographed as
described for 5a above to give 456 mg of a yellow oil.
After extraction into acid and basi®cation, recovery and
rechromatography yielded 424 mg (53%) of a yellow oil
For Km and Vmax determinations, the concentration of
acetylthiocholine iodide ranged from 8.52 to 93 mM.
The analogues to be tested were added to a ®nal con-
centration of 0.5 mM. In the majority of cases these
were soluble in aqueous solutions; however, where
compounds were insoluble, either dimethylsulfoxide
(DMSO) or citric acid was employed as a vehicle and
this was included always in the control.
ꢂ
1
that crystallized on standing; mp 88±90 C. H NMR
(CDCl3) d 2.05 (m, 1H), 2.48 (s, 3H), 2.40±2.48 (m, 3H),
3.65 (d, J=9.46 Hz, 1H), 3.70 (m, 1H), 4.24 (d,
J=17.45 Hz, 1H), 4.92 (d, J=17.45 Hz, 1H), 7.16 (d,
J=6.85 Hz, 1H), 7.25 (m, 2H), 7.53 (d, J=7.22 Hz, 1H).
(The presence of cis isomer (10%) was indicated by a
doublet of the benzylic hydrogen at d 3.45 (J=2.63 Hz).
IR (neat ®lm): 3430, 3345, 1670, 1655, 1575, 1283, 1093,
1080, 975, 749 cm 1; MS (FAB): m/z 217 (M+H, 100),
186 (41), 154 (15), 133 (27), 118 (14); HRMS (FAB)
calcd for C13H17N2O (M+H): 217.1341; obs. 217.1359.
Anal. calcd for C13H16N2O: C, 72.19; H, 7.46; N, 12.95.
Found: C, 72.11; H, 7.33; N, 12.88.
Passive avoidance training. Postnatal day 80 male Wis-
tar rats (300±350 g) were obtained from the Biomedical
Facility, University College Dublin. These were housed
singly in a 12 h light/dark cycle with food and water
available ad libitum. Animals employed for neurobeha-
vioural studies were maintained and handled in the test
environment for 5 days prior to the commencement of
studies. Animals were handled daily for a minimum of 3
min and, on days 3, 4 and 5, their weights monitored
and spontaneous behavior was assessed in an open ®eld
apparatus for 5 min. On day of training and immedi-
ately preceding time of recall, spontaneous behavior was
reassessed. All studies were carried out in a sound-
proofed, darkened room and all experimental proce-
dures were approved by the Review Committee of the
Biomedical Facility of University College, Dublin and
were carried out by individuals who held the appro-
priate licence issued by the Ministry of Health.
cis-10-N-Methylaminobenzindolizidin-3-one (29a).
A
mixture of imine 28 (159 mg) and Pd/C (10%, 110 mg)
in EtOH (10 mL) was hydrogenated (50 psi) overnight,
and then ®ltered. The ®ltrate was concentrated in vacuo
and the residue was chromatographed on silica gel elut-
ing with (CH2Cl2:MeOH:NH4OH (150:8:1) to give 29a
1
as an oil (11 mg, 8%): H NMR (300 MHz) d 7.25 (m,
4H), 4.92 (d, J=18.0 Hz, 1H), 4.31 (d, J=17.7 Hz, 1H),
3.95 (m, 1H), 3.43 (d, J=2.7 Hz, 1H), 2.38 (s, NCH3);
MS (FAB), m/e 217 (100, M+H), 154 (55), 136 (45).
Animals were trained in a one-trial, step-through, light±
dark passive avoidance paradigm as described pre-
viously.18 The apparatus consisted of a box measuring
300 mm wide Â260 mm deepÂ270 mm high. The front
and top were transparent, allowing observation of
behavior inside the apparatus. The box was divided into
two compartments, separated by a central shutter that
contained a small opening 50 mm wide and 75 mm high.
The smaller of the compartments measured 90 mm in
width and contained a low power (6v) illumination
sourceÐthe lighted compartment. The larger compart-
ment measured 210 mm in width and was not illumi-
nated. The ¯oor of the dark compartment consisted of
a grid of stainless steel bars which could deliver a
Determination of cholinesterase and acetylcholinesterase
activities. Wistar rat brain (postnatal day 80) was
homogenized in 10% (w/v) of icecold 50 mM Tris-buf-
fered saline (0.9% w/v) containing ethylenediaminete-
traacetic acid (1 mM), bovine serum albumin (1% w/v)
and Triton X 100 (0.5% w/v). Homogenates were cen-
trifuged at 10,000 g for 10 min at 4 ꢂC. Supernatants
were decanted and stored in aliquots at 70 ꢂC until
required. Protein content was determined by the
method of Lowry et al.15
Enzyme activity was assayed according to the spec-
trophotometric method of Ellman et al.16 with the

1646
S. Zhao et al. / Bioorg. Med. Chem. 7 (1999) 1637±1646
remotely-controlled, scrambled footshock (0.75 mA
every 0.5 ms) of 5 s duration. Training entailed the ani-
mal being placed within the light compartment, and
recording the latency to enter the dark compartment
with all four paws; at that time the animal received the
remote footshock. Animals were tested for recall of this
inhibitory stimulus prior to sacri®ce by placing them
into the light compartment and noting their latency to
enter the dark compartment. A criterion period of 600 s
was used. On the day of training animals were adminis-
tered the test compound or saline control via the intra-
peritoneal route 30 min prior to training. Animals
rendered amnesic received scopolamine (0.8 mg/kg ip)
20 min prior to training.
P.; Brùgger, S.; Kofod, H., Eds.; Munksgaard: Copenhagen,
1992; p 354.
3. (a) Inoue, A.; Kawai, T.; Wakita, M.; Iimura, Y.; Sugimoto,
H.; Kawakami, Y. J. Med. Chem. 1996, 39, 4460. (b) Kawa-
kami, Y.; Inoue, A.; Kawai, T.; Wakita, M.; Sugimoto, H.;
Hop®nger, A. J. Bioorg. Med. Chem. 1996, 4, 1429. (c) Hey-
dorn, W. E. Exp. Opin. Invest. Drugs 1997, 6, 1527.
4. Drugs Future 1994, 19, 656; 1996, 21, 766.
5. Manning, F. C. American Family Physician 1994, 50, 819.
6. (a) Osterrieder, W. Br. J. Pharmac. 1987, 92, 521; Edwards,
G.; Weston, E. H. In Receptor Data for Biological Experi-
ments; Ellis Horwood: New York, 1991; p 194. (b) Evans, J.
M.; Hamilton, T. C.; Longman, S. D.; Stemp, G.; Eds.;
Potassium Channels and their Modulators: From Synthesis to
Clinical Experience; Taylor and Francis, 1996. (c) Glover, W.
E. Gen. Pharmacol. 1982, 13, 259. (d) Drugs Future 1995, 20,
142.
Recall latency was assessed at 24 h post training. The
results are expressed as the mean‹SEM and sig-
ni®cance determined using the Mann±Whitney U-test
for nonparametric data.
7. O'Neill, P. M.; Bray, P. G.; Hawley, S. R.; Ward, S. A.;
Park, B. K. Pharmacol. Ther. 1998, 77, 29.
8. Drugs Future 1991, 16, 33; 1994, 19, 343.
9. Gonzalez, Trigo G.; Alvarez-Builla, J. An Quim Ser C 1980,
76, 12: CA, 94, 139584c.
10. Rausser, R.; Weber, L.; Hershberg, E. B.; Oliveto, E. P. J.
Org. Chem. 1966, 31, 1342.
11. Bolton, R.; Danks, T. N.; Paul, J. M. Tetrahedron Lett.
1994, 35, 3411.
Acknowledgements
12. Chaki, H.; Yamabe, H.; Sugano, M.; Morita, S.; Bessho,
T.; Tabata, R.; Saito, K.-I.; Egawa, M.; Tobe, A.; Morinaka,
Y. Bioorg. Med. Chem. Lett. 1995, 5, 1489.
This research was supported in part by American Bio-
genetic Sciences.
13. Chaki, H.; Yamabe, H.; Sugano, M.; Morita, S.; Bessho,
T.; Tabata, R.; Saito, K.-I.; Egawa, M.; Tobe, A.; Morinaka,
Y. Bioorg. Med. Chem. Lett. 1995, 5, 1495.
References
14. (a) Rigo, B.; Kolocouris, N. J. Heterocyclic Chem. 1983,
51, 893. (b) Martin, L. L.; Scott, S. J.; Agnew, M. N.; Setes-
cak, L. L. J. Org. Chem. 1986, 51, 3697.
15. Lowry, O. H.; Rosenbrough, N. J.; Farr, A. L.; Randall,
R. J. J. Biol. Chem. 1995, 193, 265.
16. Ellman, G. L.; Courtney, K. D.; Andres, V., Jr.; Feath-
erstone, R. M. Biochem. Pharmacol. 1961, 7, 88.
17. Whittaker, M. Methods of Enzyme Analysis 1984, 4, 52.
18. Fox, G. B.; O'Connell, A. W.; Murphy, K. J.; Regan, C.
M. J. Neurochem. 1995, 65, 2796.
1. (a) Cacabelos, R.; Nordberg, A.; Caamano, J.; Franco-
Maside, A.; Fernandez-Novoa, L.; Gomez, M. J.; Alvarez, X.
A.; Takedas, M.; Prous, J., Jr.; Nishimura, T.; Winblad, B.
Drugs of Today 1994, 30, 295. (b) Allain, H., Bentue-Ferrer, D.;
Belliard, S.; Derouesne, C. Prog. Med. Chem. 1997, 34, 46. (c)
Crimson, M. L. Ann. Pharmacother. 1994, 28, 744. (d) Her-
shenson, F. M.; Moos, W. H. J. Med. Chem. 1986, 29, 1125.
2. Hershenson, F. M. In New Leads and Targets in Drug
Research, Alfred Benzon Symposium 33; Krogsgaard-Larsen,