Syntheses and evaluation of halogenated cytisine derivatives and of bioisosteric thiocytisine as potent and selective nAChR ligands

Article abstract of DOI:10.1016/S0223-5234(01)01222-3

We have developed one-step syntheses of halogenated derivatives of (-)-cytisine featuring a halogen substituent at positions 3, 5 or 3 and 5 of the 2-pyridone fragment, and prepared the novel bioisosteric thiocytisine by oxygen-sulphur exchange. The affinities of these pyridone-modified analogs of (-)-cytisine for (α4)₂(β2)₃ and α7* nAChRs in rat forebrain membranes were determined by competition with (±)-[³H]epibatidine and [³H]MLA, respectively. The 3-halocytisines 7 possess subnanomolar affinities for (α4)₂(β2)₃ nAChRs, higher than those found for (-)-cytisine as well as for the 5-halocytisines 8 and 3,5-dihalocytisines 6. In contrast to the parent alkaloid the 3-halogenated species display much a higher affinity for the α7* nAChR subtype. The most potent molecule was 3-bromocytisine (7b) with preferential selectivity (200-fold) for the (α4)₂(β2)₃ subtype [K_i = 10 pM (α4β2) and 2.0 nM (α7*)]. Replacement of the lactam with a thiolactam pharmacophore to thiocytisine (12) resulted in a subnanomolar affinity for the (α4)₂(β2)₃ nAChR subtype (K_i = 0.832 nM), but in a drastic decrease of affinity for the α7* subtype; thiocytisine (12) has a K_i value of 4000 nM (α7*), giving a selectivity of 4800-fold for the neuronal (α4)₂(β2)₃-nAChR and thus displaying the best affinity-selectivity profile in the series under consideration.

Full text of DOI:10.1016/S0223-5234(01)01222-3

Eur. J. Med. Chem. 36 (2001) 375–388
´
© 2001 Editions scientifiques et me´dicales Elsevier SAS. All rights reserved
PII: S0223-5234(01)01222-3/SCO
Laboratory Note
Syntheses and evaluation of halogenated cytisine derivatives and of bioisosteric
thiocytisine as potent and selective nAChR ligands
Peter Imminga, Paul Klaperskia, Milton T. Stubbsa, Gunther Seitza*, Daniela Gu¨ndischb
aDepartment of Pharmaceutical Chemistry, Philipps-University, Marbacher Weg 6, D-35032 Marburg, Germany
bDepartment of Pharmaceutical Chemistry, Rheinische Friedrich-Wilhelms-University, Kreuzbergweg 26,
D-53115 Bonn, Germany
Received 17 October 2000; revised 19 January 2001; accepted 1 February 2001
Abstract – We have developed one-step syntheses of halogenated derivatives of (−)-cytisine featuring a halogen substituent at
positions 3, 5 or 3 and 5 of the 2-pyridone fragment, and prepared the novel bioisosteric thiocytisine by oxygen–sulphur exchange.
The affinities of these pyridone-modified analogs of (−)-cytisine for (a4)2(b2)3 and a7* nAChRs in rat forebrain membranes were
determined by competition with ( )-[3H]epibatidine and [3H]MLA, respectively. The 3-halocytisines 7 possess subnanomolar
affinities for (a4)2(b2)3 nAChRs, higher than those found for (−)-cytisine as well as for the 5-halocytisines 8 and 3,5-dihalocytisines
6. In contrast to the parent alkaloid the 3-halogenated species display much a higher affinity for the a7* nAChR subtype. The most
potent molecule was 3-bromocytisine (7b) with preferential selectivity (200-fold) for the (a4)2(b2)3 subtype [Ki=10 pM (a4b2) and
2.0 nM (a7*)]. Replacement of the lactam with a thiolactam pharmacophore to thiocytisine (12) resulted in a subnanomolar affinity
for the (a4)2(b2)3 nAChR subtype (Ki=0.832 nM), but in a drastic decrease of affinity for the a7* subtype; thiocytisine (12) has a
Ki value of 4000 nM (a7*), giving a selectivity of 4800-fold for the neuronal (a4)2(b2)3-nAChR and thus displaying the best
´
affinity–selectivity profile in the series under consideration. © 2001 Editions scientifiques et me´dicales Elsevier SAS
halocytisines / thiocytisines / nAChR subtype selectivity / bioisosterism
1. Introduction
to nicotine addiction, tardive dyskinesia and
schizophrenia. Moreover, worldwide interest in
nAChR agonists as potential analgesics has emerged,
representing an attractive area of research in pain
control [1, 2, 4, 6, 11, 14]. Although several promising
ligands for nAChRs have been developed [1–18] dur-
ing the past few years, there is still a need for potent
agents that would interact more selectively with the
neuronal nicotinic receptors and display no or mini-
mal side-effects as compared with naturally occurring
prototypical agonists for nAChRs such as (−)-
nicotine (1) or (−)-epibatidine (2) [1, 4, 10, 18], whose
therapeutic usefulness is limited by several untoward
effects. (see figure 1.)
There is accumulating evidence that ligands acting
with high agonistic affinity at nicotinic acetylcholine
receptors (nAChRs) may possibly be utilised as thera-
peutics [1–18] in the treatment of various neurologi-
cal and mental disorders related to a decrease in
cholinergic function: these include, e.g. senile demen-
tia of the Alzheimer type, Parkinson’s disease, atten-
tion deficit hyperactivity disorder, Tourette’s syn-
drome, depression and ulcerative colitis, in addition
Abbreviations: nAChR, nicotinic acetylcholine receptor; MLA,
methyllycaconitine; SAR, structure–activity relationship; PET,
positron emission tomography; DMAP, 4-dimethylaminopyridine;
BTX, bungarotoxine; HEPES, N-(2-hydroxyethyl)-piperazine-N%-(2-
ethanesulphonic acid); HSS, HEPES–salt solution; HRMS, high-reso-
lution mass spectrometry.
1,2,3,4,5,6-Hexahydro-1,5-methano-pyrido[1,2-a][1,
5]diazocin-8-one, (−)-cytisine (3) [19, 20] with num-
bering through the text, easily accessible by extracting
seeds from Laburnum anagyroides medicus (Fabaceae)
[21], showed a remarkable combination of properties,
* Corresponding author
E-mail address: seitzg@mailer.uni-marburg.de (G. Seitz).

376
P. Imming et al. / European Journal of Medicinal Chemistry 36 (2001) 375–388
in medicinal chemistry, retaining activity both in vitro
and in vivo [32], we anticipated that an oxygen–sul-
phur exchange in the pharmacophoric lactam moiety
might lead to a retention of activity or even enhanced
potency associated with higher selectivity. This would
reveal whether the size of the substituents and the
presence of hydrogen bond acceptors were important
factors for the affinity of the nAChR ligand. Such
compounds might also possess reduced toxicity.
A recent paper describing the synthesis of analogs
of (−)-cytisine for in vivo studies of nAChRs using
PET [33] (positron emission tomography) and several
patents [18, 34] claiming (−)-cytisine (3) and its
derivatives as useful agents for the treatment of neu-
rogenerative diseases and their use in addiction ther-
apy prompted us to report our own independently
obtained results.
In this communication we describe one-step prepa-
rations of several halogenated cytisine derivatives and
the first synthesis of the previously unknown thio-
cytisine, as well as the in vitro binding affinities of the
target molecules for (a4)2(b2)3 and a7* nAChRs, de-
termined by competition with ( )-[3H]epibatidine [35]
and [3H]MLA, respectively [36].
sharing various physiological effects with (−)-nicotine
(1) [22]. It is acutely toxic and its toxicological re-
sponses include nausea, convulsions, and death by
respiratory failure. In comparison with (−)-nicotine
(1), however, (−)-cytisine (3) is a more potent nAChR
ligand, displaying higher selectivity toward the
(a4)2(b2)3 nAChR subtype combined with sub-
nanomolar affinity [22–25]. Advantageously the
Laburnum-alkaloid shows a longer half-life in vivo
than (−)-nicotine [26], crosses the blood–brain-bar-
rier [18] and exhibits comparable efficacy and potency
to (−)-nicotine in stimulating dopamine release from
striatal synaptosomes [18].
Until now, only a few structure–activity relation-
ship (SAR) studies on (−)-cytisine derivatives have
been reported [18, 27]. Thus, in our effort to gain
novel nAChR agonists [28, 29] with possibly im-
proved pharmacodynamic profiles and safety over the
natural alkaloids 1, 2 and 3, we decided to synthesise
a number of structurally modified derivatives of (−)-
cytisine. Replacement of a hydrogen atom by a halo-
gen in one or more specific positions of a biologically
active lead compound may substantially improve the
intrinsic biological activity [15, 30, 31]. Thus, we
investigated the influence of halogen substituents such
as chlorine, bromine and iodine on the in vitro
affinity for (a4)2(b2)3 and a7* nAChRs, in order to
gain further insight into the SAR of these species.
In addition we were interested in the divalent
bioisosteric replacement [32] of the lactam pharma-
cophore, the hydrogen bond acceptor functionality of
which is required for maintaining biological activity
[4, 23]. Since bioisosteric replacement of a lactam by a
thiolactam pharmacophore has been used extensively
2. Chemistry
(−)-Cytisine (3) is a chiral quinolizidine alkaloid
with a tricyclic skeleton, consisting of the A-, B-, and
C-rings. It is characterised by a bispidine framework 4
fused to a 2-pyridone moiety. The constitution of 3
has been elucidated by chemical degradation [37] and
by syntheses [38–42]; the absolute configuration of
the two chiral centers was established by Okuda et al.
to be 7R,9S [37]. The solution and crystal structure of
the alkaloid 3 has been studied using NMR tech-
niques and X-ray crystallography, respectively [43].
The 2-pyridone moiety of 3 allows various elec-
trophilic substitution reactions [19, 33, 34]. These
properties inspired us to prepare primarily the halo-
genated cytisines 6–8 (compounds with the a designa-
tion are chloro derivatives, whereas the b and c
designations denote, respectively, bromo and iodo
compounds).
Although a variety of synthetic approaches to halo-
cytisines has been developed [19, 33, 34, 44], most of
the previously known halogenation reactions of the
alkaloid 3 afforded three-step procedures (including
N-12 protecting and deprotecting besides the halo-
Figure 1. Natural ligands for nAChRs and the bispidine core
of (−)-cytisine (3).

P. Imming et al. / European Journal of Medicinal Chemistry 36 (2001) 375–388
377
reagent was employed, a mixture of three halogenated
products was obtained with the monosubstituted spe-
cies 7 and 8 predominating. The mixture of the three
halocytisines of types 6, 7 and 8 could be easily
separated by careful flash chromatography on silica
gel, and converted subsequently into the crystalline
hydrogen fumaric salts 9–11. The structures of the
crystalline salts 9–11 were conclusively confirmed by
1
mass, H and 13C NMR spectral analysis (see Section
6).
Since bioisosterism serves as a valuable aid in SAR
studies and new drug design [32], we were interested
in the previously unknown bioisosteric thioanalog of
(−)-cytisine herein named thiocytisine. It was antici-
pated that (−)-cytisine (3) might easily be converted
to the corresponding thioanalog by treatment with
convenient sulphur transfer reagents. Because the well
known thionating agent tetraphosphorus decasul-
phide (for a recent review see Ref. [45]) involved a
long reaction time combined with low yields, we
utilised a novel, microwave-accelerated, only recently
published method [46]. In this process (−)-cytisine (3)
was simply mixed with Lawesson’s reagent (0.5
equiv.) and then irradiated under solvent free condi-
tions, to obtain the desired (−)-thiocytisine (12) with
nearly 40% yield.
Figure 2. Reagents and conditions: (1) Hal=Cl; N-chlorosuc-
cinimide (1 equiv. mol) in AcOH/H2O=3:2, 1 h reflux,
column chromatography (silica gel, CHCl3/CH3OH=10:1);
fraction 1, 6a (5%); fraction 2, 8a (40%); fraction 3, 7a (26%).
(2) Hal=Br; N-bromosuccinimide (1 equiv. mol) in AcOH/
H2O=3:2, 2 h reflux, column chromatography (silica gel,
CHCl3/CH3OH=10:1); fraction 1, 6b (5%); fraction 2, 8b
(27%); fraction 3, 7b (27%). (3) Hal=I; iodine chloride (1
equiv. mol) in AcOH (100%), 16 h r.t., column chromatogra-
phy (silica gel, CHCl3/CH3OH=10:1); fraction 1, 6c (1%);
fraction 2, mixture 7c, 8c (51%); fraction 3, 8c (10%).
Detailed spectroscopic data analysis (MS, IR, UV,
1H and 13C NMR) conclusively proved the expected
structure of 12. The mass spectrum of 12 shows three
signals in the region of the molecular ion (m/z=208,
M+ +2, 207, M+ +1, and 206, M+), a typical pattern of
a compound containing one sulphur atom. Substitu-
tion of sulphur for oxygen shifted the UV absorbance
maximum from umax=320 nm for (−)-cytisine (3) to
umax=357 nm for (−)-thiocytisine (12). The IR spec-
trum of 12 in the solid state (KBr) exhibited a strong
absorption band (w CꢀS) at 1206 cm−1, providing
evidence for the presence of the thiocarbonyl group
[47–50], while the lactam 3 displayed only compara-
tively weak absorptions in this region. (A similar
strong absorption is also found in the acetylated
thiocytisine (13).) (See figures 3 and 4.)
genation step). Thus, we employed a technique that
proved effective for a one-step procedure. The general
route for the cytisine derivatives is illustrated in figure
2. The key step in the one-step syntheses of the
halocytisines 6-8 was the halogenation reaction of the
N-12 protonated (−)-cytisinium acetate (5), prepared
in situ by employing aqueous acetic acid (60%) as the
solvent and utilising N-chloro-, N-bromosuccinimide
and iodine chloride (ICl) as halogen transfer reagents.
Using a manifold excess of the halogenating agent, a
twofold substitution of the 3- and 5- positions of the
2-pyridone moiety occurred, leading nearly quantita-
tively to the halocytisines of type 6. However, when
only one molar equivalent of the halogen transfer
The 1H and 13C NMR spectra of solutions of
(−)-thiocytisine (12) in CHCl3-d1, taken at 400 and
100.5 MHz, respectively are similar to those of the
parent alkaloid [43] (see Section 6). The thiocarbonyl
carbon resonance was observed at l=179.82, as-
signed on the basis of its chemical shift. Eventually,
appropriate crystals for an X-ray crystallographic
analysis were obtained, which conclusively established

378
P. Imming et al. / European Journal of Medicinal Chemistry 36 (2001) 375–388
(−)
the substituents to generate a d charge on S, it is
this resonance that increases the charge density on the
thiocarbonyl sulphur and makes hydrogen bond for-
mation possible (normally the sulphur atom in a
thione group is a much weaker hydrogen bond accep-
tor than the oxygen atom of a CꢀO group) [51].
Acetylation of the secondary amine 3 with an ex-
cess of acetyl chloride/DMAP was successful in
providing the acetylated derivative 13 (as a pair of
the constitution of compound 12 and revealed the
intact bispidine structure annulated with a 2-thiopyri-
done moiety.
Figure 4 shows the molecular structure of 12 (with
the atomic numbering used) and the conformation
that (−)-thiocytisine (12) adopts in the solid state.
The A ring of 12, as in (−)-cytisine (3) [43], is planar,
and the bond length and angles give evidence of a
considerable delocalisation of electrons on the ring.
Ring B exhibits a half-chair, and ring C a rigid chair
conformation with C-7, C-9, C-11, and C-13 lying in
1
rotamers about the NꢁCꢀO bond). The H NMR
spectrum of 13 was recognised as a series of doubled
signals, stemming from two rotamers [52].
,
a plane. The C(2)ꢀS bond length of 1.684(5) A is
,
much longer (0.074 A) than that in thioketones, (with
3. Pharmacology
,
a value of 1.61 A as a reasonable baseline [50, 51])
due to the resonance in the thiolactam group. Because
The cytisine derivatives listed in table I were tested
for their in vitro affinity for (a4)2(b2)3 and a7*
nAChRs subtypes by radioligand binding assays. To
determine the affinities for the (a4)2(b2)3 nAChR sub-
type a previously described competition assay was
used with ( )-[3H]epibatidine and the P2 membrane
fraction of Sprague–Dawley rat forebrain. These
studies demonstrated that the specific binding of ( )-
[3H]epibatidine to crude synaptic membranes of rat
forebrain, at concentrations up to 800 pM, is charac-
terised by a single population of binding sites with
Kd =8 0.3 pM [35]. It has been previously found
that the predominant receptor with high affinity for
[3H]nicotine, (−)-[3H]cytisine, ( )-[3H]epibatidine,
and 5-[125I]iodo-A-85380 in rat brain is composed of
a4 and b2 subunits [31]. As shown in table I, competi-
tion assays yielded Ki values of (−)-cytisine for the
(a4)2(b2)3 nAChR subtype that are consistent with
recently reported in vitro measurements of the natural
alkaloid [35] using the same assay conditions.
CꢀS bonds appear to rely on the (+)-M-influence of
Figure 3. Reagents and conditions: (1) Lawesson’s reagent,
microwave irradiation (800 W), alumina column chromatogra-
phy, 37%. (2) DMAP, Na2CO3, CH3COCl; 67%.
To characterise the binding of each of the (−)-
cytisine variants to the a7* nAChR subtype, [3H]-
MLA and membrane fractions isolated from the rat
brain were used. [3H]MLA bound to a single popula-
tion of binding sites exhibited a Kd value of 1.2 0.2
nM (n=3). The affinity determined was in good
agreement with previously published values [36]
(Kd =1.86 nM). [3H]MLA binds to rat brain mem-
branes with a regional distribution characteristic of
a-BTX-sensitive, putative a7* subunit-containing
nAChRs [36]. The parent compound (−)-cytisine
showed a Ki value of 261 nM using [3H]MLA. This
pattern agrees with previous findings [31] where Ki
values (260 20 nM) for (−)-cytisine were determined
using[125I]a-BTX.
Figure 4. ortep diagram (50% probability ellipsoids) showing
the solid state conformation of 12. The hydrogen atoms are
drawn as spheres with arbitrary radii.

P. Imming et al. / European Journal of Medicinal Chemistry 36 (2001) 375–388
379
Table I. Radioligand binding affinities of several cytisine variants to (a4)2(b2)3 and a7* nAChRs in comparison with (−)-nicotine,
(9)-epibatidine and (−)-cytisine.

380
P. Imming et al. / European Journal of Medicinal Chemistry 36 (2001) 375–388
As shown in table I, competition assays yielded Ki
affinity than the parent alkaloid 3 (Ki =0.122 nM),
similar to ( )-epibatidine (2) (Ki =0.008 nM). The
halogen substituents at position 5 of the 2-pyridone
moiety exerted an evidently smaller impact on the
binding affinity, with Ki values dropping in the higher
picomolar (Br) and even nanomolar ranges (Cl). Two-
fold halogenation in the 3,5-position of the 2-pyri-
done moiety significantly reduced the affinity. Thus,
the emerging trend, that the presence of a bulkier
substituent at position 5 of the 2-pyridone pharma-
cophore reduces the affinity for the (a4)2(b2)3 nAChR
subtype, is corroborated with the 3,5-dihalocytisines
6a–c [Ki values range from 0.520 (I) to 10.8 nM (Br)].
The affinities of caulophylline methiodide and
caulophylline for the (a4)2(b2)3 receptor exceeded
their affinities for the a7* subtype much more than
that observed with the parent (−)-cytisine (3). Re-
markably, the 3-halocytisines exhibited the highest
affinity for the a7* subtype in this series, comparable
with the affinities of a-BTX and MLA. Halogenation
in position 5 or twofold halogenation in 3 and 5
positions resulted in a decrease of affinity for the a7*
subtype compared with those of the 3-halocytisines.
The divalent bioisosteric replacement of the lactam
oxygen in 3 by sulphur yielded (−)-thiocytisine (12)
values of 0.838 nM for (−)-nicotine (1) and 0.008 nM
for ( )-epibatidine (2). These results are consistent
with recently reported in vitro measurements of the
natural alkaloids [15].
4. Results and discussion
We have demonstrated that halogenation of the
2-pyridone moiety of (−)-cytisine (3) can be carried
out successfully in a one-step approach utilising (−)-
cytisinium acetate (5), in situ prepared in aqueous
acetic acid (60%) as the starting material. The synthe-
ses of the twofold substituted halocytisines of type 6
were accomplished with yields better than 80%, em-
ploying a twofold excess of the halogenating agent
mentioned. The same method proved useful in the
preparation of the 3-halocytisines 7 and the 5-halo-
cytisines 8 utilising one molar equivalent of the halo-
gen transfer reagent. The resulting mixture of the
halocytisines 6, 7 and 8 could be separated by simple
chromatography on silica gel. The ratio of the 3- and
5- regioisomers was shown to be dependent on the
halogen transfer reagent used (ratios of isolated yield
of 6/7/8: a, 5:40:26%; b, 5:27:27%; c, 1:35:19%). An
efficient comparison of our results with the literature
data proved difficult because detailed information
concerning the yields obtained with the three-step
procedures (including an additional chromatographic
separation) in the patent [34] and the short communi-
cation [33] were not available in all cases. Neverthe-
less the one-step procedure described herein, in our
opinion, seemed to be advantageous.
Compared with ( )-epibatidine (2), (−)-cytisine (3)
exhibited a 15-fold lower affinity (Ki =0.122 nM) for
the (a4)2(b2)3 subtype and a 60-fold lower affinity for
the a7* subtype (Ki =261 nM). N-methylation of 3 to
caulophylline caused a dramatic loss in affinity drop-
ping into the nanomolar range (Ki=5.7 nM) for
(a4)2(b2)3 nAChRs, whereas the dimethylated salt,
caulophylline methiodide disclosed only a twofold
lower affinity (Ki=0.238 nM) in comparison with the
parent alkaloid.
characterised by
a
thiolactam pharmacophore.
Though somewhat less potent, the thiolactam 12
showed a sevenfold lower affinity compared with the
parent alkaloid with subnanomolar binding affinity
for the (a4)2(b2)3 receptor (Ki =0.832 nM). Remark-
ably, the novel (−)-thiocytisine shows the best
affinity–selectivity profile for (a4)2(b2)3 nAChRs in
this series with an affinity for the a7* subtype Ki =
4000 nM. The observed differences in affinity of (−)-
cytisine (3) and (−)-thiocytisine (12) agree well with
the Sheridan model [53] of the nicotinic pharma-
cophore, suggesting that the optimal distances (a–
,
,
,
b=4.7 A; a–c=4.0 A; b–c=1.2 A) between three
pharmacophoric elements are most important for the
three-dimensional arrangement also for (−)-cytisine
(3) and its derivatives, recognised by the nAChRs:
1. a cationic center [e.g. the basic or quaternised
N-12 atom of ring C of (−)-cytisine (3)]
2. an electronegative atom capable of accepting a
hydrogen bond (e.g. the lactam carbonyl oxygen
of 3)
3. a dummy point or an atom to define a line along
which the hydrogen bond may form (in the case of
3 the lactam carbonyl carbon).
The introduction of halogen substituents in most
cases resulted in a retention of affinity. The 3-halo-
cytisines retained subnanomolar affinity (Ki values of
0.01–0.022 nM) to central (a4)2(b2)3-nAChRS (Br>
I>Cl>H); the bromo representative of this series 7b
showed even an about one order of magnitude higher
A detailed analysis of the structure and conforma-
tion of (−)-cytisine (3) in the solution and in the solid

P. Imming et al. / European Journal of Medicinal Chemistry 36 (2001) 375–388
381
state [43] revealed that the alkaloid featured similar
spatial arrangements of the pharmacophoric ele-
ments (a) and (b) to (−)-nicotine (1): the N-12ꢁO-2
order of magnitude and comparable to the affinity of
( )-epibatidine (2). In contrast, (−)-cytisine deriva-
tives of type 8 with a 5-halo-substituted 2-pyridone
ring exhibit a substantially lower affinity similar to
the corresponding 3,5-dihalocytisines 6. The thiolac-
tam 12 was found to retain much of the biological
activity of the parent alkaloid with a subnanomolar
binding affinity for the (a4)2(b2)3 nAChR subtype.
For the a7* nAChR subtype affinities, the 3-halo-
cytisines proved the most active species in the series
under consideration, comparable with the affinities
of a-BTX and MLA. Bioisosteric CꢀO/CꢀS replace-
ment resulted in a drastic decrease of affinity for the
a7* subtype: thus, the novel thiocytisine (12) displays
the best affinity–selectivity profile for (a4)2(b2)3
nAChRs in this series. It would be very interesting
to find out in future studies what kind of substitu-
tion in the (−)-cytisine template would be able to
change the selectivity toward the a7* receptor.
,
distance of 4.89 A agrees well with the current nico-
tinic pharmacophor model (calculated internitrogen
,
distance in 1: 4.87 A) [53]. As expected, the divalent
bioisosteric replacement of the lactam by the thiolac-
tam pharmacophore changes the three-dimensional
arrangement of the ligand [32]. The probable reasons
for the lower affinity of (−)-thiocytisine (12) are,
firstly, the different a–b distance (N-12ꢁS-2) of 5.248
,
A (surprisingly similar to the calculated internitrogen
distance in ( )-epibatidine (2)) and the a–c distance
,
(N-12ꢁC-2) of 4.290 A, taken from the X-ray data of
12 (see Section 6), and secondly, the CꢀO/CꢀS ex-
change resulting in a modification of the hydrogen-
bond acceptor ability, crucial for nAChR affinity
[23, 53]. As already discussed the effective elec-
tronegativity of S in thiolactams is increased by con-
jugative interactions between CꢀS and the lone pair
of the vicinal N substituent, a clear example of reso-
nance-induced hydrogen bonding at sulphur accep-
6. Experimental protocols
tors
[51].
Comparing
the
frequencies
of
hydrogen-bond formation of the CꢀS and CꢀO ac-
ceptors (e.g. (−)-cytisine (3) and (−)-thiocytisine
(12)) the CꢀS···HN hydrogen bonds are anticipated
to be significantly weaker than their CꢀO···HN
analogs, thus accounting for the lower (a4)2(b2)3
affinity of 12 and the dramatic loss of affinity for the
a7* nAChR subtype.
6.1. General chemistry
Standard vacuum techniques were used in the hand-
ling of the air-sensitive materials. Melting points were
determined on a ‘Leitz-Heiztischmikroskop’ HM-Lux
and are uncorrected. Solvents were dried and freshly
distilled before use according to literature procedures.
IR spectra were recorded on a Perkin–Elmer 257, 398
and a Nicolet FT-IR spectrometer 510-P; liquids were
run as films, solids as KBr pellets. H NMR and 13C
1
5. Conclusions
NMR were recorded on a JEOL JNM-GX 400 and LA
500 and l values are given in ppm relative to te-
tramethylsilane as internal standard (J values in Hz).
Mass spectra were measured with a Fisons Instruments
VG 70-70 E spectrometer at a 70 eV ionising voltage
(EI). Column chromatography was carried out on
Merck silica gel 40 (40–60 mesh, flash chromatogra-
phy) or Merck silica 60, 70–230 mesh. Reactions were
monitored by thin-layer chromatography (TLC) by us-
ing plates of silica gel (0.063–0.200 mm, Merck) or
silicagel-60F254 microcards (Riedel de Haen). Optical
rotations were determined on a Jasco DIP-370 polar-
imeter. UV–vis spectra were recorded on a Shimadzu
scanning spectrophotometer UV-2101 PC. Combustion
analyses were performed internally. Microwave irradia-
tion was carried out in a domestic oven (800 W) privi-
leg 8017.
In our continuing efforts to develop new ligands
with high affinity and better selectivity for the multi-
various subtypes of nAChRs, we have prepared and
evaluated several analogs of the highly toxic alkaloid
(−)-cytisine (3). We found that several of these (−)-
cytisine-based compounds with halogen atoms as the
substituents at position 3, 5, or 3 and 5 of the 2-pyri-
done fragment or characterised by a bioisosteric thi-
olactam pharmacophore instead of the lactam
functionality proved to be highly potent nAChR lig-
ands. Most of these species possess subnanomolar
affinity for the (a4)2(b2)3 nAChR subtype in mem-
branes from rat forebrain. In this series, 3-bromo-
cytisine (7b) shows the highest affinity for (a4)2(b2)3
nAChR subtype (Ki =10 pM), exceeding that of the
parent alkaloid (−)-cytisine (3) by approximately one

382
P. Imming et al. / European Journal of Medicinal Chemistry 36 (2001) 375–388
6.2. Synthetic procedures
viscous oil. The crude material was purified by silica gel
chromatography (21×3 cm, CHCl3–CH3OH=10:1),
obtaining three fractions. Isolated yields were as follows:
fraction 1 (6a), colourless, viscous oil, 35.6 mg (5%),
fraction 2 (7a), yellowish viscous oil, 270.0 mg (40%),
fraction 3 (8a), yellow crystalline powder, 175.8 mg
(26%).
6.2.1. (−)-3,5-Dichlorocytisine
(9,11-dichloro-1,2,3,4,5,6-hexahydro-1,5-methano-
pyrido-[1,2-a](1,5-)diazocin-8-one) (6a) [33, 34, 44]
To a stirred solution of 3 (190.2 mg, 1.00 mmol) in
aqueous acetic acid (10 mL, 60%) was added dropwise
at reflux temperature a solution of N-chlorosuccinimide
(280.2 mg, 2.10 mmol) in aqueous acetic acid (10 mL,
60%) within 10 min. After stirring for 1.5 h at reflux
temperature the solvent was removed in vacuo at 70°C
(bath temperature). The residue was treated with
aqueous NaOH (10 mL, 10%) and the resulting aqueous
solution extracted four times with CHCl3 (10 mL). The
combined organic phase was dried over anhydrous
MgSO4 (3 g), filtered, and the solvent removed in vacuo
at 45°C (bath temperature), to give 304 mg of a viscous
yellow oil. The crude material was purified by silica gel
chromatography (18.5 cm×3 cm, CHCl3/CH3OH=5:1)
to yield 6a (249.8 mg, 83%) as pale yellow, viscous oil,
Rf=0.38 (silica gel, CHCl3/CH3OH=5:1). [h]2D0= −
38.6 (c=0.3, CH3OH); IR (KBr): w (cm−1) 3364 (NH),
1648 (CO). UV (CH3OH): umax (log m)=213 nm (3.96),
6.2.3. (−)-3-Chlorocytisine
(9-chloro-1,2,3,4,5,6-hexahydro-1,5-methano-pyrido[1.2-a]
[1.5]-diazocin-8-one) fumarate (10a)
Using the conditions employed for the preparation of
11a, 7a (92.7 mg, 0.41 mmol) was treated with fumaric
acid (48.5 mg, 0.41 mmol) to yield the salt 10a (114.3
mg, 81%), colourless crystals, m.p. 196–198°C. [h]2D0= −
81.5 (c=0.2, CH3OH); IR (KBr): w (cm−1) 3428 (NH),
1660 (CO). UV (CH3OH): umax (log m=211 nm (4.17),
1
236 (3.84), 320 (3.91). H NMR (DMSO-d6, 400 MHz):
l=1.87 (m, 1H, 8-Ha), 1.90 (m, 1H, 8-Hb), 2.37 (s,
2
3
broad, 1H, 9-H), 2.91 (dd, J=11.9 Hz, J=6.2 Hz,
1H, 11-Ha), 2.93 (dd, 2J=11.7 Hz, 3J=6.0 Hz, 1H,
13-Ha), 2.99 (d, J=11.7 Hz, 1H, 13-Hb), 3.05 (m, 2H,
2
2
3
11-Hb, 7-H), 3.80 (dd, J=15.4 Hz, J=6.6 Hz, 1H,
10-Ha), 3.94 (d, 2J=15.4 Hz, 1H, 10-Hb), 6.11 (d,
3J=7.7 Hz, 1H, 5-H), 6.54 (s, 2H, vinyl-H), 7.62 (d,
3J=7.5 Hz, 1H, 4%-H). 13C NMR (DMSO-d6, 125.8
MHz): l=24.5 (C-8), 26.0 (C-9), 33.3 (C-7), 49.9 (C-
10), 50.5 (C-11), 51.5 (C-13), 103.7 (C-5), 120.1 (C-3),
134.2 (2 vinyl-C), 136.9 (C-4), 149.8 (C-6), 158.3 (C-2),
166.5 (2 carboxylate-C). Anal. Found: C, 52.74; H, 5.04;
N, 8.19. Calc. for C15H17ClN2O5 (341.37): C, 52.87; H,
5.03; N, 8.22%.
1
244 (3.75), 332 (3.85). H NMR (CHCl3-d1, 400 MHz):
l=1.91 (m, 1H, 8-Ha), 2.00 (m, 2H, 8-Hb, NH), 2.38 (s,
2
3
broad, 1H, 9-H), 2.96 (dd, J=11.9 Hz, J=2.2 Hz,
2
1H, 11-Ha), 3.00 (d, J=12.3 Hz, 1H, 13-Ha), 3.10 (d,
2J=12.3 Hz, 1H, 13-Hb), 3.16 (d, 2J=11.9 Hz, 1H,
2
11-Hb), 3.40 (s, broad, 1H, 7-H), 3.97 (dd, J=15.7 Hz,
3J=5.7 Hz, 1H, 10-Ha), 4.16 (d, 2J=15.7 Hz, 1H,
1
10-Hb), 7.57 (s, 1H, 4-H). (For the H NMR (250 MHz)
see Ref. [34]. 13C NMR (CHCl3-d1, 100.5 MHz): l=
26.0 (C-8), 27.2 (C-9), 32.3 (C-7), 50.1 (C-10), 51.8
(C-11)1, 52.5 (C-13)1, 108.9 (C-5), 122.6 (C-3), 137.8
(C-4), 145.9 (C-6), 158.5 (C-2). MS (70 eV, 130°C):
m/z(%)=260 (20, M++2), 259 (5, M++1), 258 (35, M+),
44 (100). HRMS. Found: 258.0280. Calc. for
C11H12Cl2N2O: 258.0326.
6.2.4. (−)-5-Chlorocytisine (11-chloro-1,2,3,4,5,
6-hexahydro-1,5-methano-pyrido[1.2-a][1.5]-
diazocin-8-one) fumarate (11a)
To a stirred solution of compound 8a (131.2 mg, 0.58
mmol) in 2-propanol (2 mL) was added at reflux tem-
perature fumaric acid (68.1 mg, 0.58 mmol) in 2-
propanol (3 mL). The stirred solution was kept at 22°C
for 4 h, the precipitated solid was collected, washed
three times with anhydrous diethyl ether (5 mL) and
dried over silica in vacuo, yielding colourless crystals
(174.7 mg, 88%), m.p. 188–190°C. [h]2D0= −36.5 (c=
0.1, CH3OH); IR (KBr): w (cm−1) 3444 (NH), 1667
(CO), 1654 (CO). UV (CH3OH): umax (log m)=209 nm
6.2.2. Chlorination of (−)-cytisine 3 with 1 equiv. mol. of
N-chlorosuccinimide to 6a, 7a and 8a
To a stirred solution of 3 (570.9 mg, 3.00 mmol) in
aqueous acetic acid (10 mL, 60%) was added dropwise
at reflux temperature a solution of N-chlorosuccinimide
(401.0 mg, 3.00 mmol) in aqueous acetic acid (10 mL,
60%) within 8 min. After stirring for 1 h at reflux
temperature the solvent was removed in vacuo at 70°C
(bath temperature) to give 1517 mg of a yellowish,
1
(4.27), 239 (4.00), 322 (3.80). H NMR (DMSO-d6, 400
MHz): l=1.85 (m, 1H, 8-Ha), 1.92 (m, 1H, 8-Hb), 2.35
2
3
(s, broad, 1H, 9-H), 2.88 (dd, J=11.8 Hz, J=1.8 Hz,
1 Assignment not confirmed.
1H, 11-Ha), 2.91 (dd, 2J=12.2 Hz, 3J=2.5 Hz, IH,

P. Imming et al. / European Journal of Medicinal Chemistry 36 (2001) 375–388
383
13-Ha), 3.01 (m, 2H, 13-Hb, 11-Hb), 3.31 (s, broad, 1H,
3J=6.7 Hz, 1H, 10-Ha), 4.14 (d, 2J=15.6 Hz, 1H,
10-Hb), 7.87 (s, 1H, 4-H). *Assignment not confirmed.
For the H NMR spectrum of the corresponding hydro-
2
3
7-H), 3.74 (dd, J=15.7 Hz, J=6.6 Hz, 1H, 10-Ha),
2
3
1
3.85 (d, J=15.4 Hz, 1H, 10-Hb), 6.29 (d, J=9.5 Hz,
1H, 3-H), 6.56 (s, 2H, vinyl-H), 7.43 (d, 3J=9.7 Hz, 1H,
4-H). 13C NMR (DMSO-d6, 125.8 MHz): l=24.8 (C-8),
25.9 (C-9), 31.2 (C-7), 48.5 (C-10), 49.8 (C-11), 50.7
(C-13), 108.3 (C-5), 116.6 (C-3), 134.0 (2 vinyl-C), 139.4
(C-4), 146.8 (C-6), 160.9 (C-2), 166.3 (2 carboxylate-C).
Anal. Found: C, 52.82; H, 5.18; N, 8.12. Calc. for
C15H17ClN2O5 (341.37): C, 52.87; H, 5.03; N, 8.22%.
chloride salt see Ref. [34]. 13C NMR (CHCl3-d1, 100.5
MHz): l=26.1 (C-8), 27.3 (C-9), 34.7 (C-7), 50.1 (C-10),
52.2 (C-13), 52.5 (C-11), 97.3 (C-5), 112.51 (C-3), 143.6
(C-4), 147.9 (C-6), 158.7 (C-2). MS (70 eV, 160°C): m/z
(%)=350 (46, 81Br, 81Br−M+), 349 (13, 79Br, 81Br−M++
1), 348 (91, 79Br, 81Br−M+), 347 (8, 79Br, 79Br−M++1),
346 (47, 79Br, 79Br−M+), 305 (100). Anal. Found: C,
37.90; H, 3.63; N, 8.01. Calc. for C11H12Br2N2O (348.0):
C, 37.96; H, 3.47; N, 8.05%.
6.2.5. (−)-3,5-Dibromocytisine
(9,11-dibromo-1,2,3,4,5,6-hexahydro-1,5-methano-
6.2.6. Bromination of (−)-cytisine 3 with 1 equiv. mol.
of N-bromosuccinimide to 6b, 7b and 8b
pyrido-[1.2-a][1.5]diazocin-8-one) (6b) [33, 34, 44, 54]
To a stirred solution of 3 (190.2 mg, 1.00 mmol) in
acetic acid (100%, 25 mL) was added dropwise at room
temperature a solution of bromine (0.20 mL, 622 mg,
3.91 mmol) in acetic acid (100%, 3 mL). After stirring for
3 h ice water (100 mL) was added and the mixture stirred
vigorously for 30 min. The resulting suspension was
filtered, and the collected solid washed twice with diethyl
ether (10 mL). After drying the orange solid (394 mg) in
vacuo it was suspended in a mixture of ethanol (6 mL)
and water (4 mL). The stirred suspension was heated to
reflux for 5 min and the resulting solution cooled to ca.
5°C. The crystalline precipitate was collected, washed
twice with diethyl ether (10 mL) and dried in vacuo over
P4O10 to give the hydrobromide of 6b as colourless
needles, m.p. 220–226°C (dec.) Lit. [54]: m.p. 238°C.
(−)-3,5-Dibromocytisine-hydrobromide (179 mg, 0.42
mmol) was dissolved in CH2Cl2 (10 mL). To the solution
was added water (20 mL) and aqueous sodium hydroxide
(2 mL, 10%). After vigorous stirring of the two-phases-
system, CH2Cl2 (10 mL) was added, the aqueous layer
separated, and extracted three times with CH2Cl2 (30
mL). The combined organic phases were dried over
Na2SO4 (3 g), filtered, and the solvent was removed in
vacuo. The crude material was purified by silica gel
chromatography (10 cm×2 cm, CHCl3/CH3OH=5:1) to
yield 6b (134 mg, 92%), which appeared to be a colourless
solid, m.p. 108–110°C (Lit. [54]: m.p. 67–75°C). [h]2D0 (of
the corresponding fumarate salt)= −23.1(0.2, MeOH);
IR (KBr): w (cm−1) 3319 (NH), 1642 (CO). UV
(CH3OH): umax (log m)=222 nm (3.97), 244 (3.82), 334
(3.96). 1H NMR (CHCl3-d1, 400 MHz): l=1.80 (s,
broad, 1H, NH), 1.93 (m, 1H, 8-Ha), 1.97 (m, 1H, 8-Hb),
To a stirred solution of 3 (951 mg, 5.00 mmol) in
aqueous acetic acid (20 mL, 60%) was added dropwise at
reflux temperature a solution of N-bromosuccinimide
(890.4 mg, 5.00 mmol) in aqueous acetic acid (30 mL,
60%) within 15 min. After stirring for 2 h at reflux
temperature the solvent was removed in vacuo at 75°C
(bath temperature) to yield a brown, viscous oil (2960
mg), which was purified by silica gel chromatography
(20×3 cm, CHCl3–CH3OH=10:1), obtaining three frac-
tions. The isolated yields were as follows: fraction 1 (6b),
90.8 mg (5%), colourless solid, m.p. 108–110°C, fraction
2 (7b), 362.3 mg (27%), colourless solid, m.p. 51–53°C,
fraction 3 (8b), 364.9 mg (27%), pale yellow solid, m.p.
103–106°C.
6.2.7. (−)-3-Bromocytisine (9-bromo-
1,2,3,4,5,6-hexahydro-1,5-methano-pyrido[1.2-a][1.5]-
diazocin-8-one) (7b)
[h]2D0= −83.3 (c=0.2, CH3OH); IR (KBr): w (cm−1
)
3429 (NH), 1641 (CO). UV (CH3OH): umax (log m)=210
1
nm (3.94), 239 (3.63), 323 (3.90). H NMR (CHCl3-d1,
400 MHz,): l=1.75 (s, broad, 1H, NH), 1.95 (m, 2H,
8-Ha, 8-Hb), 2.35 (s, broad, 1H, 9-H), 2.92 (s, broad, 1H,
7-H), 3.00 (d, 2J=12.2 Hz, 1H, 11-Ha), 3.05 (d, 2J=11.2
2
Hz, 1H, 13-13-Ha), 3.07 (d, J=11.2 Hz, 1H, H-13b),
3.11 (d, 2J=12.2 Hz, 1H, 11-Hb), 3.95 (dd, 2J=15.6 Hz,
3J=6.6 Hz, 1H, 10-Ha), 4.18 (d, 2J=15.6 Hz, 1H,
3
3
10-Hb), 5.93 (d, J=7.6 Hz, 1H, 5-H), 7.68 (d, J=7.6
Hz, 1H, 4-H). 13C NMR (CHCl3-d1, 100.5 MHz): l=
26.1 (C-8), 27.6 (C-9), 35.4 (C-7), 51.1 (C-11), 52.8
(C-13), 53.72 (C-10), 104.8 (C-5), 112.0 (C-3), 140.6
(C-4), 151.0 (C-6), 159.6 (C-2). For the 1H and 13C NMR
2
3
2.35 (m, 1H, 9-H), 2.93* (dd, J=12.0 Hz, J=2.2 Hz,
1H, 11-Ha)1, 2.99 (d, J=12.0 Hz, 1H, 11-Hb), 3.09 (d,
spectra of the hydrochloride salt see Ref. [34]. HRMS.
2
2
79
13
2J=11.9 Hz, 1H, 13-Hb), 3.18* (d, J=11.9 Hz, 1H,
Found 268.0219. Calc. for C11H BrN2O: 268.0211.
13-Ha), 3.34 (m, 1H, 7-H), 3.96 (dd, 2J=15.6 Hz,
Found: 270.0191. Calc. for C11H BrN2O: 270.0190.
81
13

384
P. Imming et al. / European Journal of Medicinal Chemistry 36 (2001) 375–388
6.2.8. (−)-5-Bromocytisin (11-bromo-
6.2.10. Iodination of(−)-cytisine 3 with 1 equiv. mol of
iodine monochloride to 6c, 7c and 8c
1,2,3,4,5,6-hexahydro-1,5-methano-pyrido[1.2-a][1.5]-
diazocin-8-on) (8b)
To a stirred solution of 3 (570.7 mg, 3.00 mmol) in
acetic acid (15 mL, 100%) was added dropwise a solution
of iodine monochloride (487 mg, 3.00 mmol) in acetic
acid (4.7 mL, 100%) at room temperature within 8 min.
The mixture was stirred for 6 h and then water (20 mL)
was added. After stirring for a further 16 h the solvent
was removed in vacuo at 75–80°C (bath temperature) to
give a yellow-brown, viscous oil (1198 mg). The crude
material was dissolved in CHCl3 (50 mL) and to the
solution was added sodium carbonate (2 g). After 1 h, the
suspension was filtered, and the solvent was removed
under reduced pressure to yield a residue (5 mL), which
was purified by silica gel column chromatography (18×3
cm, CHCl3/CH3OH=10:1), to give three fractions. Iso-
lated yields were as follows: fraction 1 (6c), 12.2 mg
(0.9%) as yellowish crystals, fraction 2, a mixture of the
compounds 8c and 7c (483 mg, 51%) as a yellowish,
viscous oil, fraction 3, (7c), as a yellow viscous oil (97 mg,
10%). Fraction 2 was purified again by silica gel column
chromatography (20×2 cm, CHCl3/CH3OH=40:1) to
give two fractions. Compounds 8c (fraction 1%) and 7c
(fraction 2%) appeared to be a colourless viscous oil (177
mg, 19%) and a yellow viscous oil (236 mg, 25%),
respectively.
[h]D20= −10.9 (c=0.2, CH3OH); IR (KBr): w (cm−1)=
3442 (NH), 1650 (CO). UV(CH3OH): umax (log m)=207
1
nm (3.94), 239 (3.76), 322 (3.66). H NMR (CHCl3-d1,
400 MHz): l=1.62 (s, broad, 1H, NH), 1.97 (m, 2H,
8-Ha, 8-Hb), 2.32 (s, broad, 1H, 9-H), 2.93 (dd, 2J=12.0
3
2
Hz, J=2.2 Hz, 1H, 11-Ha), 2.99 (d, J=12.2 Hz, 1H,
13-Ha), 3.08 (d, 2J=12.2 Hz, 1H, 13-Hb), 3.19 (d,
2J=12.2 Hz, 1H, 11-Hb), 3.34 (s, broad, 1H, 7-H), 3.91
(dd, 2J=15.6 Hz, 3J=6.6 Hz, 1H, 10-Ha), 4.08 (d,
2J=15.6 Hz, 1H, 10-Hb), 6.37 (d, 3J=9.7 Hz, 1H, 3-H),
7.42 (d, 3J=9.5 Hz, 1H, 4-H). For the 1H NMR
spectrum of the corresponding hydrochloride salt, see
Ref. [34]. 13C NMR (CHCl3-d1, 100.5 MHz): l=26.3
(C-8), 27.3 (C-9), 34.7 (C-7), 50.2 (C-10), 50.8 (C-11),
52.6 (C-13), 98.5 (C-5), 117.6 (C-3), 142.4 (C-4), 148.0
(C-6), 162.4 (C-2). HRMS. Found: 268.0187. Calc. for
79
13
81
C11H BrN2O: 268.0211. Found: 270.0177. Calc. for
C11H BrN2O: 270.0190.
13
6.2.9. (−)-3, 5-Diiodocytisine
(9,11-diiodo-1,2,3,4,5,6-hexahydro-1,5-methano-pyrido-
[1.2-a][1.5] diazocin-8-one) hydrochloride (9c)
To a stirred solution of 3 (190.2 mg, 1.00 mmol) in
acetic acid (100%, 10 mL) was added dropwise a solution
of iodine monochloride (1.38 g, 8.50 mmol) in acetic acid
(100%, 10 mL) at room temperature within 8 min. After
stirring for 4 h the solution was diluted with ice water (20
mL) and stirred vigorously for 16 h at room temperature.
The crystalline precipitate was collected on a filter,
washed twice with diethyl ether (5 mL), dried (P4O10) and
recrystalised from ethanol (10 mL) to give compound 9c
as orange crystals (470 mg, 98%), m.p. 198–202°C (dec.).
[h]2D0= −11.2 (c=0.2, CH3OH); IR (KBr): w (cm−1) 3432
(NH), 1610 (CO). UV (CH3OH): umax (log m)=206 nm
6.2.11. (−)-3-Iodocytisine (9-iodo-1,2,3,4,5,
6-hexahydro-1,5-methano-pyrido[1.2-a][1.5]-
diazocin-8-one) fumarate (10c)
A stirred solution of 7c (75 mg, 0.24 mmol) in 2-
propanol (2 mL) was heated to reflux and an excess of
fumaric acid (41.8 mg, 0.35 mmol) dissolved in 2-
propanol (3 mL) was added. The solution was kept at
room temperature for 5 h, the precipitated solid was
collected, washed three times with anhydrous diethyl
ether (5 mL) and dried in vacuo over silica, yielding
cream-coloured crystals (60.1 mg, 58%), m.p. 204–
207°C. [h]2D0 = −71.2 (c=0.2, CH3OH); IR (KBr): w
(cm−1) 3440 (NH), 2946, 1653 (CO). UV (CH3OH): umax
(log m)=210 nm (4.30), 327 (4.03). 1H NMR (DMSO-d6,
500 MHz): l=1.85 (m, 2H, 8-Ha, 8-Hb), 2.32 (s, broad,
lH, 9-H), 2.86 (d, 2J=13.5 Hz, 1H, 11-Ha), 2.89 (d,
2J=14.5 Hz, 13-Ha), 2.94 (dd, 2J=11.8 Hz, 3J=2.7 Hz,
1H, 13-Hb), 3.00 (m, 2H, 11-Hb, 7-H), 3.78 (dd, 2J=15.3
Hz, 3J=6.4 Hz, lH, 10-Ha), 3.89 (d, 2J=15.4 Hz,
10-Hb), 5.95 (d, 3J=7.3 Hz, lH, 5-H), 6.57 (s, 2H,
vinyl-H), 7.97 (d, 3J=7.3 Hz, 1H, 4-H). 13C NMR
(DMSO-d6, 125.8 MHz): l=24.7 (C-8), 26.6 (C-9), 33.6
(C-7), 50.5 (C-10), 51.1 (C-11), 52.1 (C-13), 86.7 (C-3),
1
(4.05), 240 (4.03), 339 (3.73). H NMR (DMSO-d6, 400
MHz): l=1.97 (m, 1H, 8-Ha), 2.04 (m, 1H, 8-Hb), 2.59
(s, broad, 9-H), 3.23 (m, 2H, 13a-H, 11a-H), 3.33 (m, 2H,
13b-H, 11b-H), 3.41 (s, broad, 1H, 7-H), 3.87 (dd,
3
2
2J=15.2 Hz, J=6.4 Hz, lH, 10a-H), 3.98 (d, J=15.7
Hz, 1H, 10b-H), 8.12 (s, broad, lH, NH), 8.34 (s, 1H,
4-H), 8.93 (s, broad, 1H, NH). 13C NMR (DMSO-d6,
100.5 MHz): l=22.9 (C-8), 24.6 (C-9), 35.1 (C-7), 45.6*
(C-10)1, 47.4* (C-13), 50.5 (C-11), 70.9 (C-5), 91.0 (C-3),
146.7 (C-6), 154.4 (C-4), 159.2 (C-2). *Assignments not
confirmed. MS. (70 eV) m/z (%): 443 (8, M+−Cl), 442
(100). Anal. Found: C, 27.40; H, 2.71; N, 5.90. Calc. for
C11H13C1I2N2O (478.49): C, 27.61; H, 2.74; N, 5.85.

P. Imming et al. / European Journal of Medicinal Chemistry 36 (2001) 375–388
385
2
3
105.7 (C-5), 134.0 (2 vinyl-C), 147.1 (C-4), 152.2 (C-6),
159.2 (C-2), 166.2 (2 carboxylate-C). Anal. Found: C,
41.52; H, 4.16; N, 6.48. Calc. for C15H17IN2O5 (432.84):
C, 41.68; H, 3.96; N, 6.48%.
3J=1.8 Hz, 1H, 13-Ha), 3.10 (dd, J=7.1 Hz, J=1.6
2
Hz, 1H, 13-Hb), 3.12 (d, J=11.9 Hz, 1H, 11-Hb), 4.26
(dd, 2J=16.3 Hz, 3J=6.6 Hz, 1H, 10-Ha), 4.63 (d,
3
4
2J=16.1 Hz, 1H, 10-Hb), 6.52 (dd, J=7.1 Hz, J=1.6
3
3
Hz, 1H, 5-H), 7.17 (dd, J=8.5 Hz, J=7.1 Hz, 1H,
3
4
6.2.12. (−)-5-Iodocytisine (11-iodo-1,2,3,4,5,6-
hexahydro-1,5-methano-pyrido[1.2-a][1.5]
-diazocin-8-one) fumarate (11c)
4-H), 7.69 (dd, J=8.6 Hz, J=1.5 Hz, 1H, 3-H). 13C
NMR (CHCl3-d1, 100.5 MHz): l=25.8 (C-8), 28.8 (C-
9), 36.6 (C-7), 52.8 (C-13), 53.8 (C-11), 58.1 (C-10),
113.6 (C-5), 133.3 (C-4), 133.5 (C-3), 154.3 (C-6), 179.8
(C-2). MS (70 eV) m/z (%): 208 (5, M++2), 207 (14,
M++1), 206 (100, M+). HRMS. Found: 206.0879. Calc.
for C11H14N2S: 206.0877.
The compound 11a was obtained by the same proce-
dure as described for l0c from 8c (22.7 mg, 0.07 mmol)
and fumaric acid (12.7 mg, 0.10 mmol). Yield 12.4 mg
(40%) colourless crystals, m.p. 149–152°C (dec.). [h]2D0=
−2.4 (c=0.2, CH3OH); IR (KBr): V (cm−1) 3447 (NH),
2941 (CH), 1653 (CO). UV (CH3OH): umax (log m=209
6.2.14. (−)-12-Acetyl-thiocytisine
1
nm (3.82), 237 (3.56), 324 (3.23). H NMR (DMSO-d6,
(3-acetyl-1,2,3,4,5,6-hexahydro-1,5-methano-pyrido-
500 MHz): l=1.88 (s, broad, 2H, 8-Ha, 8-Hb), 2.30 (s,
[1.2-a][1.5]diazocin-8-thione) (13)
broad, 1H, 9-H), 2.86 (d, 2J=11.9 Hz, 2H, 11-Ha,
To a stirred solution of 12 (50.9 mg, 0.25 mmol) in
CH2Cl2 (8 mL) was added Na2CO3 (212 mg, 2.00
mmol), 4-dimethylaminopyridine (10 mg, 0.08 mmol)
and dropwise a solution of acetylchloride (0.02 mL, 0.28
mmol) in CH2Cl2 (2 mL). After stirring for 16 h at room
temperature the suspension was filtered, the residue
washed three times with CH2Cl2 (10 mL), and the
resulting solution was concentrated in vacuo to yield
72.5 mg of a semicrystalline solid. The product was
purified by silica gel chromatography (20.5 cm×3 cm,
CHCl3/CH3OH=10:1) to give compound 13 as yellow
crystals (41 mg, 67%), m.p. 153–155°C. [h]2D0= −522.9
(c=0.1, CH3OH); IR (KBr): w (cm−1)=2915 (CH),
1640 (CO), 1611, 1544, 1206 (C=S). UV (CH3OH):
2
3
13-Ha), 2.99 (dd, J=12.1 Hz, J=2.7 Hz, 13-Hb), 3.02
2
(d, J=12.1 Hz, lH, 11-Hb), 3.15 (s, broad, 1H, 7-H),
2
3
3.75 (dd, J=15.5 Hz, J=6.5 Hz, 1H, 10-Ha), 3.82 (d,
2J=15.3 Hz, lH, 10-Hb), 6.08 (d, J=9.4 Hz, 1H, 3-H),
3
6.58 (s, 2H, vinyl-H), 7.64 (d, J=9.6 Hz, 1H, 4-H). 13C
3
NMR (DMSO-d6, 125.8 MHz): l=25.7 (C-8), 26.6
(C-9), 38.1 (C-7), 49.1 (C-10), 50.4 (C-11)1, 51.2 (C-13)1,
69.2 (C-5), 117.8 (C-3), 134.1(2 vinyl-C), 147.3 (C-4),
150.1 (C-6), 161.7 (C-2), 166.2 (2 carboxylate-C). Anal.
Found: C, 41.42; H, 4.07; N, 6.26. Calc. for
C15H17IN2O5 (432.84): C, 41.68; H, 3.96; N, 6.48%.
6.2.13. (−)-Thiocytisine (1,2,3,4,5,6-hexahydro-
1
1,5-methano-pyrido[1.2-a][1.5]-diazocin-8-thione) (12)
A mixture of (−)-cytisine (3) (193 mg, 1.00 mmol) and
Lawesson’s reagent (0.50 mmol) was taken in a glass
tube and mixed thoroughly with a spatula. The glass
tube was placed in an alumina bath inside the mi-
crowave oven (800 W) and irradiated for 3 min. The
coloured crude material was dissolved in CHCl3 (50
mL), the resulting solution filtered, and the solvent
removed to give a residue (3 mL); this was adsorbed on
alumina (3 g) and purified by alumina column chro-
matography (20 cm×3 cm, CHCl3/CH3OH=40:1) af-
fording the pure compound 12; yield 78.5 mg (37%) pale
yellow crystals, m.p. 163–165°C, Rf=0.45 (silica gel,
CHCl3–CH3OH=5:1). [h]2D0= −244.7 (c=0.2, CH3-
OH); IR (KBr): w (cm−1)=3426 (NH), 1206 (C=S).
UV (CH3OH): umax (log m)=209 nm (3.88), 286 (3.92),
umax (log m)=209 nm (4.04), 288 (3.94), 359 (3.84). H
NMR (CHCl3-d1, 400 MHz, two rotamers, ratio 2:1):
l=1.74 (s, 3H, COCH3), 2.01 (s, 3H, COCH3), 2.07 (s,
broad, 4H, 8-Ha, 8%-Hb), 2.67 (s, broad, 2H, 9-H, 9%-H),
2.83 (d, 2J=13.2 Hz, 1H, 11-1H, 11-Ha), 2.89 (dd,
3
2J=13.4 Hz, J=2.0 Hz, 1H, 13-Ha), 3.24 (s, broad,
2
2H, 7-H, 7%-H), 3.42 (d, J=13.4 Hz, 1H, 13-Hb), 3.47
(dd, 2J=13.2 Hz, 3J=2.0 Hz, 1H, 11-Hb), 3.87 (d,
2J=13.0 Hz, 1H, 11-Ha), 3.98 (d, 2J=13.2 Hz, 1H,
2
3
13-Ha), 4.23 (dd, J=16.5 Hz, J=6.6 Hz, 2H, 10-Ha,
10%-Ha), 4.56 (d, 2J=16.3 Hz, 1H, 10-Hb), 4.68 (d,
2J=13.2 Hz, 1H, 13%-Hb), 4.74 (d, J=16.5 Hz, 1H,
2
3
4
10%-Hb), 4.78 (d, 1H, 11%-Hb), 6.61 (m, J=7.1 Hz, J
3
3
=1.5 Hz, 2H, 5-H, 5%-H), 7.16 (2d, J=8.6 Hz, J=7.1
3
4
Hz, 2H, 4-H, 4%-H), 7.66 (2d, J=8.6 Hz, J=1.5 Hz,
2H, 3-H, 3%-H). 13C NMR (CHCl3-d1, 100.5 MHz, two
rotamers, ratio 2:1): l 20.8 (COCH3), 21.2 (COCH3),
25.4 (C-8, C-8%), 28.2 (C-9%), 28.5 (C-9), 35.4 (C-7), 35.9
(C-7%), 47.3 (C-11%), 48.3 (C-13), 52.3 (C-11), 53.4 (C-
13%), 56.9 (C-10), 57.0 (C-10%), 113.2 (C-5%), 114.4 (C-5),
1
357 (3.85). H NMR (CDCl3, 400 MHz): l=1.54 (s,
broad, 1H, NH), 1.98 (m, 2H, 8-Ha, 8-Hb), 2.47 (s,
broad, 1H, 9-H), 2.99 (dd, 2J=9.7 Hz, 3J=1.1 Hz,
2
11-Ha), 3.03 (s, broad, 1H, 7-H), 3.05 (dd, J=7.2 Hz,

386
P. Imming et al. / European Journal of Medicinal Chemistry 36 (2001) 375–388
132.8 (C-4%), 133.5 (C-4), 134.0 (C-3), 134.4 (C-3%), 151.4
(C-6%), 151.5 (C-6), 169.5 (COCH%3), 169.7 (COCH3),
179.9 (C-2), 180.8 (C-2%). MS. (70 eV) m/z (%): 250 (6,
M++2), 249 (16, M++1), 248 (100, M+); HRMS. Found:
248.0996. Calc. for C13H16N2SO: 248.0983.
total of 925 reflections was collected. The intensities of
three representative reflections were measured after ev-
ery 150 reflections. No decay correction was applied.
The linear absorption coefficient, v, for Cu-Ka radia-
tion is 24.8 cm−1. An empirical absorption correction
based on azimuthal scans of several reflections was
applied, which resulted in transmission factors ranging
from 0.84 to 1.00. The data were corrected for Lorentz
and polarisation effects.
6.3. Crystal-structure determination
6.3.1. Preparation of single crystals for X-ray analysis
Compound 12 (10 mg) was dissolved in CHCl3 (0.5
mL) and filtered through a folded filter paper in such a
way that the filtrate directly went into a test tube (bore
ca. 1 cm). Then cyclo-hexane (3 mL), as a suitable
precipitant, was layered carefully down the side of the
tube on to the solution. The tube was then corked and
left to stand undisturbed for 12 h, furnishing yellow
crystals suitable for X-ray crystallographic analysis.
The structure was solved by direct methods [55] and
expanded using Fourier techniques [56]. The non-hydro-
gen atoms were refined anisotropically. The hydrogen
atoms were included but not refined. The final cycle of
full matrix least-squares refinement2 was based on 872
observed reflections (I>3.00|(I)) and 127 variable
parameters and converged (largest parameter shift was
0.03 times its e.d.s.) with unweighted and weighted
agreement factors of R=ꢀ ꢁꢁ Fo ꢁ−ꢁ Fc ꢁꢁ/ꢀ ꢁ Fo ꢁ=0.037;
and Rw=ꢂ(S (ꢁFoꢁ−ꢁFcꢁ)2/S F2o)=0.039. The standard
6.3.2. Crystal-structure determination of 12
‚
‚
A yellow prism crystal (ca. 0.50×0.50×0.30 mm),
obtained by recrystallisation from CHCl3/cyclohexane
by the liquid diffusion method, was mounted on a glass
fiber and investigated on a Rigaku AFC5R diffractome-
ter with graphite monochromated Cu Ka radiation and
a rotating anode generator (Rigaku). Empirical formula
C11H14N2S, molecular mass 206.30 a.u. Cell constants
and an orientation matrix for data collection, obtained
from a least-squares refinement using the setting angles
of 25 carefully centered reflections in the range 77.63<
2[<79.85° corresponded to a primitive orthorhombic
deviation of an observation of unit weight was 3.39. The
weighting figure was based on counting statistics and
included a factor (p=0.04) to downweight the intense
reflections. Plots of ꢀ ‚(ꢁ Fo ꢁ−ꢁ Fc ꢁ)2 versus ꢁFoꢁ, reflec-
tion order in data collection, sin q/u and the various
classes of indices showed no unusual trends. The maxi-
mum and minimum peaks on the final difference
−
−3
,
Fourier map corresponded to 0.17 and 0.21 e A
respectively.
,
The neutral atom scattering factors were taken from
Cromer and Waber [57]. Anomalous dispersion effects
were included in Fcalc [58]; The values for Df % and Df %
were those of Creagh and McAuley [59]. The values for
the mass attenuation coefficients are those of Creagh
and Hubbell [60]. All calculations were performed using
the teXsan [61] crystallographic software package of
Molecular Structure Corporation.
cell with dimensions a=11.271(2), b=12.622(3), c=
3
,
7.147(3) and V=1016.8(5) A . For Z=4 and F.W.=
206.30, the calculated density is 1.35 g cm−3. The
systematic absences of: h00: h"2n, 0k0: k"2n, 001:
l"2n uniquely determine the space group to be: P21212
(c19). The data were collected at a temperature of
23 1°C using the ꢀ–2[ scan technique to a maximum
2[ value of 120.5°. Omega scans of several intense
reflections, made prior to data collection, had an aver-
age width at half-height of 0.26° with a take-off angle of
6.0°. Scans of (1.37+0.30 tan [)° were made at a speed
of 32.0° min−1 (in omega). The weak reflections (I<
10.0|(I)) were rescanned (maximum of 4 scans) and the
counts were accumulated to ensure good counting statis-
tics. Stationary background counts were recorded on
each side of the reflection. The ratio of peak counting
time to background counting time was 2:1. The diameter
of the incident beam collimator was 1.0 mm, the crystal
to detector distance was 400 mm, and the detector
aperture was 9.0 mm×13.0 mm (horizontal×vertical). A
6.4. In vitro receptor binding
6.4.1. Materials
( )-[3H]Epibatidine (33 Ci mmol−1) was obtained
from NEN Life Science Products (Cologne, Germany).
[3H]MLA (20 Ci mmol−1) was purchased from Tocris/
2 Least-squares: functional minimized (ꢀ‚(ꢁFoꢁ−ꢁFcꢁ)2 where
w=1:|2c(Fo)=[|c2(Fo)+p2:4F2o]−1
, |c(do)=e.s.d. based on
counting statistics; p=p-factor. Standard deviation of an ob-
'
servation of unit weight:
(ꢀ‚(ꢁFoꢁ−ꢁFcꢁ)2/(No−Nw) where:
No=number of observations, Nw=number of variables.

P. Imming et al. / European Journal of Medicinal Chemistry 36 (2001) 375–388
387
Biotrend (Cologne, Germany). All other chemicals used
6.4.5. Data analysis
were obtained from Sigma-Aldrich (Deisenhofen, Ger-
many). Frozen Sprague–Dawley rat brains were pur-
chased from Pel-Freez Biologicals (Rogers, AR).
Competition binding data were analysed using nonlin-
ear regression methods. Ki values were calculated by the
Cheng–Prusoff [62] equation based on the measured
IC50 values and Kd=10 pM for binding of ( )-
[3H]epibatidine and Kd=1 nM for [3H]MLA. The Kd
values were obtained from five independent experiments
preformed on the same membrane preparations that
were used for the competition assays.
6.4.2. Membrane preparation
Frozen rat brains were thawed at 22°C for 30–60 min
before membrane preparation. A crude membrane frac-
tion (P2) was isolated as previously described [15, 31,
35] and stored in aliquots at −80°C for at least 16 h but
no more than 4 weeks before use. The pellets of the
crude membrane fraction were washed just once on the
day of assay. The pellets were homogenised in 30 vol-
umes of a HEPES–salt solution (HSS) containing
HEPES (15 mM), NaCl (120 mM), KCl (5.4 mM),
MgCl2 (0.8 mM), and CaCl2 (1.8 mM). After centrifuga-
tion at 35,000×g for 10 min, the resultant pellets were
resuspended in a fresh HSS and used for binding assay.
7. Supplementary material
Crystallographic data (excluding structure factors)
for the structure(s) reported in this paper have been
deposited with the Cambridge Crystallographic Data
Center as supplementary publication no. CCDC-
144569. Copies of the data can be obtained free of
charge on application to The Director, CCDC, 12
Union Road, Cambridge CB2 1EZ, UK (fax: +44-
1223-336033; e-mail: deposit@ccdc.ca-m.ac.uk).
6.4.3. (h4)2(i2)3 Binding assay
Binding assay was performed following a published
procedure [15]. Briefly, assays were carried out in HSS
at 22°C and performed in duplicate. Nonspecific binding
was determined in the presence of 300 mM (−)-nicotine.
Each assay sample of a total volume of 0.5 mL con-
tained 60 mg of membrane protein, 0.5 nM ( )-
[3H]epibatidine, and 0.2 mL of a test compound. The
samples were incubated for 90 mm, and the incubation
were terminated by vacuum filtration through Whatman
GFIB glass fiber filters, presoaked in 1% poly(ethylen-
imine) using a Brandel 48-channel cell harvester. The
radioactivity was measured using a liquid scintillation
counter (Tri-Carb 2100 TR, Packard, Dreieich,
Germany).
Acknowledgements
Financial support from Deutsche Forschungsgemein-
schaft and the Fonds der Chemischen Industrie is grate-
fully acknowledged. We thank Bayer, A.G., Merck,
A.G., and Degussa, A.G. for supplying us with valuable
chemicals.
References
[1] Lloyd G.K., Williams M.J., Pharmacol. Exp. Ther. 292 (2000)
461–467 (and the literature cited therein).
[2] Schmitt J.D., Bencherif M. (Eds.), Annual Reports in Medici-
nal Chemistry, vol. 35, Academic Press, New York, 2000, pp.
41–52.
6.4.4. h7* Binding assay
[3] Gispen W.H., Bruinwels J., Nijkamp P., Versteeg D.H.G.
The procedure used in the radioligand binding assay
was reported in detail by Davies et al. [36] with minor
modifications. Nonspecific binding was measured in the
presence of 1 mM MLA (methyllycaconitine). Each as-
say sample contained 50 mL of the test compound, 100
mL [3H]MLA to achieve a final concentration of 2 nM,
and 100 mL resuspended membranes. The samples were
incubated for 180 min at 22°C and the incubation was
terminated by rapid filtration under vacuum through
Whatman GFIB filters presoaked in 1% poly(ethylen-
imine). The radioactivity bound to the filters was mea-
sured by a liquid scintillation counter.
(Eds.), Eur. J. Pharmacol. 393 (1–3) (2000) 1–340.
[4] Holladay M.W., Dart M.J., Lynch J.K., J. Med. Chem. 40
(1997) 4169–4194.
[5] Prendergast M.A., Jackson W.J., Terry A.V. Jr., Decker M.W.,
Arneric S.P., Buccafusco J.J., Psychopharmacology 136 (1998)
50–58.
[6] Ghelardini C., Galeotti N., Barlocco D., Bartolini A., Drug
Dev. Res. 40 (1997) 251–258.
[7] Menzaghi F., McClure D.A., Lloyd G.K., in: Arneric S.P.,
Brioni J.D. (Eds.), Neuronal Nicotinic Receptors: Pharmacol-
ogy and Therapeutic Opportunities, Wiley–Liss, New York,
1998, pp. 379–394.
[8] Levin E.D., Bettergowda C., Blosser J., Gordon J., Behav.
Pharmacol. 10 (1999) 675–680.

388
P. Imming et al. / European Journal of Medicinal Chemistry 36 (2001) 375–388
[36] Davies A.R.L., Hardick D.J., Blagbrough I.S., Potter B.V.L.,
[9] McDonald I.A., Vernier J.M., Cosford N., Corey-Naeve J.,
Wolstenholme A.J., Wonnacott S., Neuropharmacology 38
(1999) 679–690.
Curr. Pharmacol. Des. 2 (1996) 357–366.
[10] Badio B., Garaffo H.M., Spande T.F., Daly J.W., Med. Chem.
[37] Okuda S., Tsuda K., Kataoka H., Chem. Ind. (London) (1961)
1751.
Res. 4 (1994) 440–448.
[11] Bai D., Xu R., Zhu X., Drugs Future 22 (1997) 1210–1220.
[38] (a) van Tamelen E.E., Baran J.S., J. Am. Chem. Soc. 77 (1955)
4944–4955. (b) van Tamelen E.E., Baran J.S., J. Am. Chem.
Soc. 78 (1956) 2913–2914. (c) van Tamelen E.E., Baran J.S., J.
Am. Chem. Soc. 60 (1958) 4659–4670.
[12] Wright E., Gallagher T., Sharples C.G.V., Wonnacott S.,
Bioorg. Med. Chem. Lett. 7 (1997) 2867–2870.
[13] Badio B., Garaffo H.M., Plummer C.V., Padgett W.L., Daly
J.W., Eur. J. Pharmacol. 321 (1997) 189–194.
[39] Bohlmann F., Englisch A., Ottawa N., Sander H., Weise W.,
Angew. Chem. 67 (1955) 708.
[14] Bannon A.W., Decker M.W., Holladay M.W., Curzon P.,
Donnelly-Roberts D., Puttfarcken P.S., Bitner R.S., Diaz A.,
Dickenson A.H., Williams M., Arneric S.P., Science 279 (1998)
77–81.
[40] Govindachari T.S., Rajadurai S., Subramanian M., Thyagara-
jan B.S., J. Chem. Soc. (1957) 3839–3844.
[41] O’Neill B.T., Yohannes D., Bundesmann M.W., Arnold E.P.,
Org. Lett. 2 (2000) 4201–4204.
[15] Koren A.O., Horti A.G., Mukhin A.G., Gu¨ndisch D., Kimes
A.S., Dannals R.F., London E.D., J. Med. Chem. 41 (1998)
3690–3698.
[42] Coe J.W., Org. Lett. 2 (2000) 4205–4208.
[43] Mascagni P., Christodoulou M., Gibbons W.A., Asres K.,
Phillipson J.D., Niccolai N., Mangani St., J. Chem. Soc. Perkin
Trans. II (1987) 1159–1165.
[16] Williams M., Sullivan J.P., Arneric S.P., Drug News Perspect.
7 (1994) 205.
[17] Perry E.K., Morris C.M., Court J.A., Cheng A., Fairbairn
A.F., McKeith I.G., Irving D., Brown A., Perry R.H., Neuro-
science 64 (1995) 385–395.
[44] Orjales A., Ribas L., Varela A., An. Quim. 68 (1972) 1419–
1432.
[45] Hartke K., Gerber H.D., J. Prakt. Chem. 338 (1996) 763–765.
[46] Varma R.S., Kumar D., Org. Lett. 1 (1999) 697–700.
[18] Lin N.-H., Meyer M.D., Exp. Opin. Ther. Patents 8 (1998)
991–1015.
[19] Leonard N.I., in: Manske R.H.F., Holmes H.L. (Eds.), The
Alkaloids, vol. 3, Academic Press, New York, 1953, pp. 119–
199.
[47] Spinner E., J. Chem. Soc. (1960) 1237–1242.
[48] Seitz G., Mann K., Schmiedel R., Matusch R., Chem. Ber. 112
(1979) 990–999.
[20] Ohmiya S., Saito K., Murakoshi I., in: Cordell G.A. (Ed.), The
Alkaloids, vol. 47, Academic Press, New York, 1995, pp.
1–114.
[49] Mayer R., Schinke E., Rosmus P., Scheithauer S., J. Prakt.
Chem. 312 (1970) 767–775.
[50] Borowiak T., Dutkiewicz G., Wysocka W., Kolanos R., Z.
Naturforsch. B 54 (1999) 1307–1312.
[21] Bojadschiewa M., Issaev I., Totev I., Dimov C., Pharmazie 26
(1971) 643.
[51] Allen F.H., Bird C.M., Rowland R.S., Raithby P.R., Acta
Crystallogr., Sect. B 53 (1997) 680–695.
[22] Anderson D.J., Arneric S.P., Eur. J. Pharmacol. 253 (1994)
261–267.
[52] Cox C., Lectka T., J. Org. Chem. 63 (1998) 2426–2427.
[23] Glennon R.A., Dukat M., Med. Chem. Res. (1996) 465–486.
[53] Sheridan R.P., Nilakantan R., Dixon J.S., Venkataraghavan
R., J. Med. Chem. 29 (1986) 899–906.
[24] Hall M., Zerbe L., Leonard S., Freedman R., Brain Res. 600
(1993) 127–133.
[54] Luputiu G., Moll F., Arch. Pharm. 304 (1971) 151–158.
[25] Schmeller T., Sauerwein M., Sporer F., Wink M., Mu¨ller W.E.,
J. Nat. Prod. 57 (1994) 1316–1319.
[55] Altomare A., Cascarano M., Giacovazzo C., Guagliardi A.,
Completion and refinement of crystal structures with SIR92, J.
Appl. Crystallogr. 26 (1993) 343–350.
[26] Sloan J.W., Martin W.R., Bostwick M., Hook R., Wala E.,
Pharmacol. Biochem. Behav. 30 (1988) 255–267.
[56] DIRDIF-94: P.T. Beurskens, G. Admiraal, G. Beurskens, W.P.
Bosman, R. de Gelder, R. Israel, J.M.M. Smits, The DIRDIF-
94 program system, Technical Report of the Crystallography
Laboratory, University of Nijmegen, The Netherlands, 1994.
[27] Boido Canu C., Sparatore F., Farmaco 54 (1999) 438–451.
[28] Che D., Siegl J., Seitz G., Tetrahedron: Asymmetry 10 (1999)
573–585.
[57] Cromer D.T., Waber J.T., International Tables for X-ray Crys-
tallography, vol. IV, The Kynoch Press, Birmingham, 1974
(Table 2.2.A).
[29] Che D., Wegge T., Stubbs M.T., Seitz G., Meier H., Methfessel
C., J. Med. Chem. 44 (2001) 47–57.
[30] Naumann K., J. Prakt. Chem. 341 (1999) 417–435.
[58] Ibers J.A., Hamilton W.C., Acta Crystallogr. 17 (1964) 781–
[31] Mukhin A.G., Gu¨ndisch D., Horti A.G., Koren A.O., Tamag-
nan G., Kimes A.S., Chambers J., Vaupel D.B., King S.,
Picciotto M.R., Innis R., London E.D., Mol. Pharm. 57 (2000)
642–649.
782.
[59] Creagh D.C., McAuley W.J., in: Wilson A.J.C. (Ed.), Interna-
tional Tables for Crystallography, vol. C, Kluwer Academic,
Boston, 1992, pp. 219–222 (Table 4.2.6.8).
[32] Patani G.A., LaVoie E.J., Chem. Rev. 96 (1996) 3147–3176.
[60] Creagh D.C., Hubbell J.H., in: Wilson A.J.C. (Ed.), Interna-
tional Tables for Crystallography, vol. C, Kluwer Academic,
Boston, 1992, pp. 200–206 (Table 4.2.4.3).
[33] Marriere E., Rouden J., Tadino V., Lasne M.C., Org. Lett. 2
(2000) 1121–1124.
[34] (a) O’Neill B.T., PCT Int. Appl. WO 98 18,798, 1998. (b)
O’Neill B.T., Chem. Abstr. 119 (1998) 4774k.
[61] teXsan: Crystal Structure Analysis Package, Molecular Struc-
ture Corporation (1985 & 1992).
[35] Gu¨ndisch D., London E.D., Terry P., Hill G.R., Mukhin A.G.,
[62] Cheng Y.-C., Prusoff W.H., Biochem. Pharmacol. 22 (1973)
3099.
Neuroreport 10 (1999) 1631–1636.
.