5-(4-Hydroxy-6-methyl-2-oxo-2H-pyran-3-yl)-7-phenyl-2,3,6,7-tetrahydro-1, 4-thiazepines as compounds with high affinity at the benzodiazepine binding site on GABAA receptors

Article abstract of DOI:10.1692/ph.2011.0732

A series of thiazepines has been studied as new ligands for the benzodiazepine binding site of the GABA_A receptor. Compounds with high affinity and weak selectivity regarding α₁β ₂γ₂, α₂β₃γ ₂, α₃β₃γ₂, and α₅β₃γ₂ subtypes were found. The pharmacophore is discussed based on experimental and theoretical results. The thiazepine sulfur atom was found to be able to act as hydrogen bond acceptor.

Full text of DOI:10.1692/ph.2011.0732

ORIGINAL ARTICLES
Biocrea GmbH, Radebeul, Germany
5-(4-Hydroxy-6-methyl-2-oxo-2H-pyran-3-yl)-7-phenyl-2,3,6,7-tetrahydro-
1,4-thiazepines as compounds with high affinity at the benzodiazepine
binding site on GABAA receptors
C. Grunwald, C. Kronbach, U. Egerland, R. Schindler, R. Dieckmann, K. Heinecke, N. Höfgen, K. Unverferth
Received August 3, 2010, accepted September 17, 2010
Dr. Christian Grunwald, Biocrea GmbH, Meissner Strasse 191,
D-01445 Radebeul, Germany
christian.grunwald@biocrea.com
Pharmazie 66: 98–104 (2011)
doi: 10.1692/ph.2011.0732
A series of thiazepines has been studied as new ligands for the benzodiazepine binding site of the GABAA
receptor. Compounds with high affinity and weak selectivity regarding ␣1␤2␥2, ␣2␤3␥2, ␣3␤3␥2, and ␣5␤3␥2
subtypes were found. The pharmacophore is discussed based on experimental and theoretical results.
The thiazepine sulfur atom was found to be able to act as hydrogen bond acceptor.
1. Introduction
Despite the difficulty in obtaining compounds with dis-
tinct affinity- or efficacy-selectivity for ␣2 or ␣3, some
efficacy-selective compounds have been synthesized and inves-
tigated. For instance, 7-[1,1-dimethylethyl)-6-(2-ethyl-2H-1,
2,4-triazol-3-ylmethoxy)-3-(2-fluorophenyl)-triazolo[4,3-b]py-
ridazine (TPA023) displays selectivity regarding ␣2/␣3 effi-
cacy. TPA023 and the very similar compound 3-(2,5-di-
fluorophenyl)-7-(1,1-dimethylethyl)-6-[(1-methyl-1 H -1,2,4,-
triazol-5-yl)methoxy]-1,2,4-triazolo[4,3-b]pyridazine (L-838,
417) do not induce sedation or ataxia in rodents or primates.
The intrinsic activities of these compounds are lower than that
GABAA receptors are the most important inhibitory transmitter
receptors in the brain. They are chloride ion channels composed
of five subunits. 21 different subunits types are known, and the
majority of GABAA channels consists of two of the six different
␣ subunits, two of the four ␤, and one of the four ␥ subunits
(Bonnert et al. 1999; Korpi et al. 1997; Mehta and Ticku 1999).
In our study, four GABAA subtypes consisting of ␣1␤2␥2,
␣2␤3␥2, ␣3␤3␥2, and ␣5␤3␥2 subunits were used. For simplicity,
we will use here just the ␣ subunit designation, e. g. ␣1, instead
of terms like “GABAA subtypes containing ␣1 subunits”.
The benzodiazepine receptor (BzR) is one of several modu-
latory binding sites at the GABAA receptor. A wide variety
of compounds are known to interact with the BzR site, mod-
ulating chloride ion conductivity of the GABAA channel and
causing, for instance, anxiolytic and anticonvulsive effects but
also unwanted side effects including sedation, ataxia, amnesia,
and dependence (Albaugh et al. 2002; Basile et al. 2004). Since
discovery of the 1,4-benzodiazepines, thousands of compounds
of different compound classes have been synthesized with the
aim to separate the desired medicinal properties from unwanted
side effects. Studies using ␣ type knockout mice suggest that
anxiolytic properties of agonists are selectively mediated by
activation of ␣2 or ␣3 subunits without inducing concomitant
sedation and memory impairment. Inverse agonists acting at the
␣5 subunit could play a role in certain aspects of cognition,
whereas agonists at ␣1 and probably also at ␣5 subunits induce
sedation (Atack 2005; Dawson et al. 2005; Möhler and Rudolph
2004; Rudolph et al. 1999; Savic et al. 2007; Sieghart 2006).
These findings prompted the synthesis of many compounds with
varying degrees of selectivity regarding affinity or efficacy for
the different GABAA subtypes over the past decade (Da Settimo
et al. 2007). However, to the best of our knowledge, no general
strategy for the design of ␣2 and ␣3 selective agonists has been
published to date. Size and topology of the included volumes
of the ␣1, ␣2, and ␣3 containing subtypes are apparently similar
(Clayton et al. 2007).
of
7-chloro-1,3-dihydro-1-methyl-5-phenyl-2H-1,4-benzo-
diazepin-2-one (diazepam) (Carling et al. 2005; Dawson
et al. 2005; Scott-Stevens et al. 2005). Thus, the discovery
of new GABAA modulators with selectivity towards ␣2 or ␣3
containing subtypes is still of interest.
2. Investigations, results and discussion
2.1. Synthesis of compounds
We discovered 5-(4-hydroxy-6-methyl-2-oxo-2H-pyran-3-
yl)-7-phenyl-2,3,6,7-tetrahydro-1,4-thiazepines as
a
new
compound class with high affinity to the benzodiazepine site
of GABAA receptors. First indications regarding the potential
utility of these compounds were found during a random high-
throughput screening of our compound library. Compounds
2-7, 9, 11, and 12 are commercially available. Compounds 1, 8,
10 were prepared according to a modified reported procedure
(Drewe et al. 2007; Sucheta et al. 1995), see Scheme 1.
For example, condensation of dehydroacetic acid with
3-chlorobenzaldehyde in ethanol and piperidine gave the corre-
sponding chalcone, which was reacted with 2-aminoethanethiol
in ethanol to produce 5-(4-hydroxy-6-methyl-2-oxo-2H-pyran-
3-yl)- 7- (3 - chlorophenyl)-2, 3, 6, 7-tetrahydro-1, 4-thiazepine,
compound 1. Enantiomers 13 and 14 were obtained by sep-
aration of compound 1 by chiral preparative HPLC. 5-(4-hy-
droxy-6-methyl-2-oxo-2H-pyran-3-yl)-7-(3-chlorophenyl)-2,3,
98
Pharmazie 66 (2011)

ORIGINAL ARTICLES
R
binding studies. IC50 values of the chemical compounds were
determined using filtration assays with 1 nM 3H-Flumazenil at
an optimized protein concentration. Diazepam was used as a
reference compound. Results are listed in Table 1.
O
OH
O
O
O
OH
O
O
O
R
1,4-thiazepines have also been studied by other groups.
Haggarty et al. (2000) have published that compounds of this
type destabilize microtubules at 20 – 50 ␮M in a colchicine-like
manner. Ohta et al. (2001) studied the inhibition of P-selectin-
dependent cell adhesion detecting compound 11 (KF38789) as
particular active with an IC50 value of 1.97 ␮M. For 7 of 10 com-
pounds under study, no or only weak cytotoxicity was found at a
concentration of 100 ␮M. Compound 11 was completely free of
cytotoxicity. Contrary to this result, 11 was found to induce apo-
ptosis by inhibiting tubulin polymerization in two human cancer
cell assays with IC50 values of 0.41 and 0.56 ␮M, respectively
(Drewe et al. 2007). Compounds 3 and 7 also displayed activity
here whereas 5, 6, 9, 10, and 12 were inactive up to 10 ␮M in
this investigation.
SH
S
NH2
H2N
H2N
OH
O
N
OH
O
N
O
NH
R
Cl
O
15
R = 3-Cl
R = 3-Cl
R = 2-Cl
R = H
1
8
10
Scheme 1: Synthesis of compounds 1, 8, 10, and 15
6,7-tetrahydro-1,4-diazepine, compound 15, was prepared
accordingtoScheme1. Thecyclizationwith1,2-diamino-ethane
was carried out on silica gel.
2.3. Structure-activity relationships
We have studied the effects of the different substituents at the
7-phenyl ring of 1,4-thiazepines on the affinity to GABAA
receptors containing ␣1, ␣2, ␣3, or ␣5 subunits, respectively.
Substituent effects apparently depend mainly on the position of
substitution (ortho, meta, para) but not on whether substituents
have polar or non-polar properties. Substitution in the meta posi-
tion enhancesaffinity forallfour subtypes. Compounds 1, 2, 3, 4,
and 5 displays higher affinity than the unsubstituted compound
10. Substitution in para position has only a marginal impact on
affinity (6, 7). The bulky isopropyl group decreases the affin-
ity towards ␣2 and ␣3 (12). Substitution in ortho position (8,
9) leaves affinities towards ␣1, ␣2, and ␣3 virtually unchanged
2.2. Biology
For our studies, human recombinant GABAA receptor subtypes
␣1␤2␥2, ␣2␤3␥2, ␣3␤3␥2, and ␣5␤3␥2 were used. The ␣1␤2␥2
subtype was used instead of ␣1␤3␥2 due to poor expression
of ␣1␤3␥2 receptors. This might reduce the comparability
of our results. However, the benzodiazepine binding site is
situated at the interface of ␣ and ␥ subunits (Hanson et al.
2008) limiting effects of this substitution. Compounds were
characterized regarding their affinity to the benzodiazepine
binding site of the 4 GABAA subtypes applying radioligand
Table 1: Influence of substituent R, ring configuration, and ring sulfur on BzR affinity
OH
N
OH
N
S
NH
Cl
R
O
O
O
O
1-14
15
a
IC [nM] or inhibition
50
b
Cpd
R
Cnf
␣ ␤ ␥
1 2
␣ ␤ ␥
2
␣ ␤ ␥
3
␣ ␤ ␥
5
3 2
2
3
2
3
2
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
dzpe
3-Cl
3-I
3-MeO
r
r
r
r
r
r
r
r
r
r
r
r
7.1 (4.0)
3.6 (0.5)
13 (2)
15 (1)
17 (2)
23 (2)
30 (8)
46 (5)
55 (10)
56 (14)
59 (2)
97 (22)
3.0 (0.4)
30 (10)
n.d.c
12 (3)
10 (2)
52 (6)
32 (2)
52 (6)
7.8 (2.5)
9.5 (1.3)
38 (11)
35 (13)
51 (1)
110 (32)
72 (17)
107 (14)
108 (14)
83 (5)
3.9 (1.4)
4.6 (2.2)
15 (4)
18 (2)
16 (5)
3,4,5-tri-MeO
3-OH
4-OH
4-Cl
2-Cl
2-NO2
H
2,4-di-MeO
4-Isopropyl
3-Cl
143 (4)
124 (27)
170 (15)
147 (12)
129 (47)
373 (44)
516 (13)
7.7 (2.2)
60 (19)
34 % (2 %)d
19 (5)
30 (5)
41 (13)
205 (24)
471 (106)
60 (7)
180 (36)
70 (6)
2.0 (0.6)
23 (9)
n.d.
189 (5)
754 (226)
4.1 (0.6)
39 (19)
(-)
(+)
r
3-Cl
45 % (3 %)d
56 (22)
a
42 (13)
35 (5)
a
Mean of IC or percentage inhibition and (standard deviation)
50
b
c
d
e
Configuration: r racemate; (-) minus enantiomer; (+) plus enantiomer; a achiral.
n.d.: not determined
Percentual inhibition at a concentration of 1000 nM
Reference diazepam
Pharmazie 66 (2011)
99

ORIGINAL ARTICLES
but reduces the ␣5 affinity. Finally, compound 11 with methoxy
substituents in ortho and para position displays a reduced affin-
ity for ␣2, ␣3, and ␣5, again compared with 10. In summary, the
investigated substituents at the 7-phenyl ring have only a weak
influence on the selectivity.
The thiazapines studied are chiral compounds. To clarify if
affinity depends on the enatiomeric form, the optically active
enantiomers 13 and 14 of the racemic mixture 1 were exam-
ined. The minus enantiomer 13 displays approximately 10-fold
greater affinity towards all receptor subtypes. As to be expected,
the affinity of the racemate lies between the affinities of the enan-
tiomers and is about half of that of the minus form. Compound
13 displays a 2-fold (␣2) up to 18-fold (␣5) higher affinity than
the reference diazepam.
Fig. 1: Hydrogen bond between thiazepine sulfur and water, optimized geometry
according to a DFT calculation. Stereo view
2.4. Pharmacophore
It is known from the pharmacophore model proposed by
Cook et al. (Clayton et al. 2007; Zhang et al. 1995) that
hydrogen bond acceptors play a significant role in binding
of BzR ligands. We have speculated that the sulfur atom of
the thiazepine ring might be involved in such an interaction.
To check this hypothesis from a theoretical point of view,
calculations based on the density functional theory (DFT) were
used to characterize the hydrogen bond acceptor nature of the
sulfur atom. In contrast to nitrogen and oxygen, sulfur is not
generally recognized for its hydrogen bond acceptor properties
in spite of theoretical (Sabin 1971) as well as experimental
findings (Adman et al. 1975; Krepps et al. 2001; Wierzejewska
2000). Recently, Wierzejewska and Sałdyka (2004, 2006) have
published results on analogous sulfur- and oxygen-containing
systems regarding their hydrogen bonding features. Using ab
initio and DFT methods, they found that complexes involving
disulfide are only slightly weaker than the corresponding
peroxide complexes. We used a modified version of the method
described by Hao (2006) to calculate the strength of hydrogen
bonds on the basis of DFT calculations. A system consisting
of 5,7-dimethyl-2,3,6,7-tetrahydro-[1,4]thiazepine and a water
molecule placed in the vicinity of the sulfur and in the direction
of one of its lone electron pairs was used and a BSSE corrected
B3LYP procedure on a 6-31++G(d,p) basis applied. An energy
of 16.849 kJ/mol and a distance of 3.35 Å were calculated for
the hydrogen bond between sulfur and water, indicating a weak
hydrogen bond. Our result agrees well with experimental data
of S···H-N bonds where distances in a range of 3.3-3.5 Å were
found (Krepps et al. 2001). The calculated geometry of the
thiazepine-water complex is depicted in Fig. 1.
To further exclude that spatial differences between the thi-
azepine ring of 1 and its diazepine analog 15 lead to the large
differences between affinities, geometries of the energetic mini-
ma were compared. They were found to be very similar with
a root mean square deviation (RMSD) of the seven-membered
rings of only 0.145 Å.
The IC50 values of compound 15 for ␣2 and ␣3 containing sub-
types are greater then 1000 nM. Compared with the most potent
enantiomer 13, 15 shows more than 130 and 240 times lower
affinity. It may therefore be assumed that the loss of affinity of
15 in comparison with the thiazepines is due to the qualitatively
changed nature of the sp3-nitrogen in place of the sulfur. Com-
pound 15 contains a secondary amine nitrogen in place of the
sulfur atom replacing the hypothetical hydrogen bond acceptor
by a hydrogen bond donor.
Fig. 2: Global energetic minimum of the R-enantiomer of compound 1. Orthographic
depiction. Non-polar and non-chiral hydrogens are not displayed
To be able to compare the geometry of our compounds with ref-
erence structures, the conformation of the seven-membered ring
of 1 was extensively analyzed. A systematic approach yielded
two chair-like, two twisted, and two boat-like conformers for
each of the two enantiomers. These 12 ring geometries were
minimized energetically using semiempirical AM1 calculations
and analyzed regarding their heats of formation. For a depiction
of the conformers and details of the calculations see the Exper-
imental section below. A flat conformation of both the R- and
the S-enantiomer was found to be energetically favored. The
global minimum is a chair-like conformation of the thiazepine
ring with the two substituents arranged in the plane of this ring
as shown in Fig. 2. The essentially flat form is consistent with
Cook’s pharmacophore (Zhang et al. 1995). As to be expected
for enantiomers, no differences were found for the two forms
besides the chirality.
Based on the overlay of 2-(4-chlorophenyl)-1H-pyrazolo[4,5-
c]quinolin-3-one (CGS-9896) and diazepam proposed by Cook
et al. (Clayton et al. 2007; Zhang et al. 1995) and our findings,
a superposition of CGS-9896, diazepam, and compound 1 was
constructed. Due to its assumed function as H-bond acceptor,
the sulfur was placed in the H1 region to enable a rough fit of
the shape of the three molecules. The chloro-phenyl ring was
aligned with the fused chloro-phenyl of diazepam and 4-chloro-
phenyl of CGS-9896 according to their lipophilic nature. The
carbonyl oxygen of the pyranone ring is able to interact with
the H-bond donor H2. The acidic OH group of the pyranone
ring corresponds to the NH group of CGS-9896 and interacts
with A2. It was not possible to favor one of the enantiomers for
the alignment as the geometry of all three structures is essen-
tially planar. The R-form was randomly selected for use here.
The resulting alignment is depicted in Fig. 3, together with the
arrangement of pharmacophoric receptor properties according
to the Cook model (Clayton et al. 2007; Zhang et al. 1995).
Taken together, experimental and computational results strongly
suggest that a H-bond acceptor is needed at the position of the
sulfur and that the sulfur atom is able to accept H-bonds. This
finding was used subsequently for an alignment of the com-
pounds under study and known BzR ligands.
2.5. Conclusion
Insummary, thiazepineshavebeenidentifiedasanewcompound
class of ligands at the benzodiazepine binding site of GABAA
100
Pharmazie 66 (2011)

ORIGINAL ARTICLES
(m, CH2-S), 3.8 and 4.2 (dd, CH2), 4.1 (m, CH2-N), 4.5 (m, CH-S), 5.7 (s,
Pyran-H), 7.4 (m, Ph-H), 13.7 (s, OH). 13C NMR (DMSO-d6) 19.6 (-CH3),
29.1 (CH2), 37.6 (CH2-N), 39.8 (CH-S), 46.4 (CH2-S), 95.9 C3 (pyran),
107.7 C5 (pyran), 127.6-130.0 (=CH-Ph), 133.3 (=C-Cl), 140.0 (=C-CH<),
162.9 (=C-CH3), 163.5 (=C-OH), 177.6 (-C=N), 183.7 O-C=O).
3.1.2.3. 5-(4-Hydroxy-6-methyl-2-oxo-2H-pyran-3-yl)-7-phenyl-2,3,6,7-
tetrahydro-1,4-thiazepine (10). Compound 10 was prepared in the same
manner as compound 1 using the appropriate chalcone (Takeuchi et al.
1980). Yield: 89%; mp: 148 ◦C. C17H17NO3S calculated m. w.: 315.39.
ES-MS: m/z = 316 (M+1). 1H NMR (DMSO-d6) 2.1 (s, -CH3), 2.9 (m,
CH2-S), 3.7 and 4.2 (dd, CH2), 4.2 (m, CH2-N), 4.5 (m, CH-S), 5.8 (s,
Pyran-H), 7.5 (m, Ph-H), 13.7 (s, OH).
13C NMR (DMSO-d6) 19.6 (-CH3), 29.4 (CH2), 39.9 (CH2-N), 41.2
(CH-S), 46.2 (CH2-S), 96.0 C3 (pyran), 107.7 C5 (pyran), 127.9-129.3
(=CH-Ph), 143.1 (=C-CH<), 162.9 (=C-CH3), 163.4 (=C-OH), 178.0
(-C=N), 183.7 O-C=O).
Fig. 3: Superposition of CGS-9896 (green), diazepam (orange), and compound 1
(white). Pharmacophoric features of the binding site according to Cook et al.
(Zhang et al. 1995) are lipophilic pockets L, sterically restricted regions S,
H-bond donors H, and an H-bond acceptor A2. Non-polar and non-chiral
hydrogens are not displayed
3.1.2.4. 5-(4-Hydroxy-6-methyl-2-oxo-2H-pyran-3-yl)-7-(3-chlorophen-
yl)-1H-2,3,6,7-tetrahydro-1,4-diazepine (15). 0.58 g (2 mmol) of 3-(3-
chloro-cinnamoyl)-4-hydroxy-6-methyl-2H-pyran-2-one (Rachedi et al.
1989), dissolved in 20 ml of ether, and 0.14 ml (2 mmol) of 1,2-diamino-
ethane, dissolved in 20 ml of ether, were mixed. 2 g of silica gel were added.
After stirring for 30 min the ether was evaporated and the residue was heated
at 80 ◦C (bath temperature) for 90 min. The silica gel was then treated with
20 ml of hot ethyl acetate and separated by filtration. The filtrate was evap-
orated to get yellow crystals. Yield: 59 %; mp: 132-134 ◦C. C17H17ClN2O3
calculated m. w.: 332.78. ES-MS: m/z = 333 (M+1). 1H NMR (DMSO-d6)
2.1 (s, -CH3), 3.1 (m, CH2-NH), 2.7 and 3.1 (dd, CH2), 3.8 (m, CH2-N,
NH), 4.5 (d, CH-N), 5.7 (s, Pyran-H), 7.4 (m, Ph-H), 13.7 (s, OH).
13C NMR (DMSO-d6) 19.1 (-CH3), 40.0 (CH2), 46.8, 46.9 (CH2-N), 57.0
(CH-N), 95.3 C3 (pyran), 107.3 C5 (pyran), 125.3-130.1 (=CH-Ph), 132.8
(=C-Cl), 147.5 (=C-CH<), 162.0 (=C-CH3), 163.3 (=C-OH), 178.7 (-C=N),
183.1 (O-C=O).
receptors. The affinity was studied with regard to substituents
at the 7-phenyl ring and configuration of the 1,4-thiazepine
ring. Potent compounds were identified. For instance, (-)-5-(4-
hydroxy-6-methyl-2-oxo-2H-pyran-3-yl)-7-(3-chlorophenyl)-
2,3,6,7-tetrahydro-1,4-thiazepine 13 displays affinities between
2 and 8 nM at ␣1␤2␥2, ␣2␤3␥2, ␣3␤3␥2, and ␣5␤3␥2 GABAA
subtypes. The investigated substituents at the 7-phenyl ring
have only weak influence on the affinity-selectivity of ␣2 and
␣3 containing subtypes compared with ␣1 and ␣5, respectively.
The overall structure of the thiazepine derivatives fits well with
an established pharmacophore. Interestingly, it was shown by
DFT calculations that the sulfur atom of the thiazepine ring is
able to act as a hydrogen bond acceptor.
3.1.3. Separation, yield, melting point, HPLC/MS, 1H- and 13C NMR
of enantiomers 13 and 14
3.1.3.1. Separation of 13 and 14. 1 g of compound 1 was stirred in 300 ml
of propan-2-ol and refluxed for 5 min under an argon atmosphere until
completely dissolved. After cooling to room temperature separation of enan-
tiomers was achieved by chiral preparative HPLC (Chiralpak AD, 50 mm x
250 mm, n-hexane/propan-2-ol = 3/2).
3. Experimental
3.1. Chemistry
3.1.1. Methods and instruments
Melting points were determined using a Boetius melting point apparatus
PHMK 05 and are uncorrected. All substances were analyzed with an Agi-
lent 1100 series HPLC/MSD system. The molecular mass was determined
using a mass selective detector after ESI in positive scan mode. Purity was
ascertained using the area percentage method on the UV trace recorded at
a wavelength of 254 nm. Purities of purchased and synthesized compounds
were found to be ≥ 95 %. Proton (1H NMR) and carbon (13C NMR) nuclear
magnetic resonance spectra were recorded on a Bruker ARX 500 NMR
spectrometer. Chemical shifts (␦) are represented in parts per million (ppm)
relative to Si(CH3)4.
3.1.3.2. (-)-5-(4-Hydroxy-6-methyl-2-oxo-2H-pyran-3-yl)-7-(3-chloroph-
enyl)-2,3,6,7-tetrahydro-1,4-thiazepine (13). The second fraction eluted
contained compound 13. Yield: 350 mg; mp: 147 ◦C; [α]D20 – 185.8◦
(EtOH). C17H16ClNO3S calculated m. w.: 349.84. ES-MS: m/z = 350
(M+1). 1H NMR (DMSO-d6) 2.1 (s, -CH3), 2.9 (m, CH2-S), 3.7 and 4.2
(dd, CH2), 4.1 (m, CH2-N), 4.5 (m, CH-S), 5.7 (s, Pyran-H), 7.4 (m, Ph-H),
13.7 (s, OH).
13C NMR (DMSO-d6) 19.6 (-CH3), 29.4 (CH2), 39.5 (CH2-N), 40.6 (CH-
S), 46.0 (CH2-S), 96.0 C3 (pyran), 107.7 C5 (pyran), 126.4-130.7 (=CH-Ph),
133.5 (=C-Cl), 145.5 (=C-CH<), 162.8 (=C-CH3), 163.5 (=C-OH), 177.6
(-C=N), 183.7 (O-C=O).
3.1.2. Preparation, yield, melting point, HPLC/MS, 1H- and 13C
NMR of compounds 1, 8, 10, 15
3.1.2.1. 5-(4-Hydroxy-6-methyl-2-oxo-2H-pyran-3-yl)-7-(3-chloropheny-
l)-2,3,6,7-tetrahydro-1,4-thiazepine (1). Compound 1 was prepared based
on a procedure published by Sucheta et al. (1995) and cited by Drewe et al.
(2007). 0.58 g (2 mmol) of the appropriate chalcone (Rachedi et al. 1989)
were dissolved in 20 ml of ether. 0.14 ml (2 mmol) of 2-aminoethanethiol
were added dropwise. 20 g of silica gel were added. After stirring for
30 min the ether was evaporated and the residue was heated at 80 ◦C (bath
temperature) for 90 min. The silica gel was then treated with 20 ml of hot
ethyl acetate and separated by filtration. The filtrate was evaporated (residue
2 – 4 ml) to get yellow crystals. Yield: 43 %; mp: 147 ◦C. C17H16ClNO3S
calculated m. w.: 349.84. ES-MS: m/z = 350 (M+1). 1H NMR (DMSO-d6)
2.1 (s, -CH3), 2.9 (m, CH2-S), 3.7 and 4.2 (dd, CH2), 4.1 (m, CH2-N),
4.2 (m, CH-S), 5.7 (s, Pyran-H), 7.4 (m, Ph-H), 13.8 (s, OH). 13C NMR
(DMSO-d6) 19.6 (-CH3), 29.4 (CH2), 39.2 (CH2-N), 40.7 (CH-S), 46.0
(CH2-S), 96.0 C3 (pyran), 107.7 C5 (pyran), 126.4-130.9 (=CH-Ph), 133.5
(=C-Cl), 145.5 (=C-CH<), 163.2 (=C-CH3), 163.5 (=C-OH), 177.6 (-C=N),
183.2 O-C=O).
3.1.3.3. (+)-5-(4-Hydroxy-6-methyl-2-oxo-2H-pyran-3-yl)-7-(3-
chlorophenyl)-2,3,6,7-tetrahydro-1,4-thiazepine (14). The first fraction
eluted contained compound 14. Yield: 250 mg; mp: 147 ◦C, [␣]D
+
20
226.8◦ (EtOH). C17
H16ClNO3S calculated m. w.: 349.84. ES-MS: m/z =
350 (M+1). 1H NMR (DMSO-d6) 2.1 (s, -CH3), 2.9 (m, CH2-S), 3.7 and
4.2 (dd, CH2),
4.1(m, CH2-N), 4.5(m, CH-S), 5.7(s, Pyran-H), 7.4(m, Ph-H), 13.7(s, OH).
13C NMR (DMSO-d6) 19.6 (-CH3), 29.4 (CH2), 39.5 (CH2-N), 40.6 (CH-
S), 46.0 (CH2-S), 96.0 C3 (pyran), 107.7 C5 (pyran), 126.4-130.7 (=CH-Ph),
133.5 (=C-Cl), 145.5 (=C-CH<), 162.8 (=C-CH3), 163.5 (=C-OH), 177.6
(-C=N), 183.7 (O-C=O).
3.1.4. HPLC/MS of commercially obtained compounds 2-7, 9, 11, 12
3.1.4.1. 5-(4-Hydroxy-6-methyl-2-oxo-2H-pyran-3-yl)-7-(3-iodophenyl)-
2,3,6,7-tetrahydro-1,4-thiazepine (2). C17H16INO3S calculated m. w.:
441.28. ES-MS: m/z = 442 (M+1).
3.1.2.2. 5-(4-Hydroxy-6-methyl-2-oxo-2H-pyran-3-yl)-7-(2-chlorophen-
yl)-2,3,6,7-tetrahydro-1,4-thiazepine (8). Compound 8 was prepared in the
same manner as compound 1 using the appropriate chalcone (Qamar and
Siddiq 1988). Yield: 36%; mp: 200 ◦C. C17H16ClNO3S calculated m. w.:
349.84. ES-MS: m/z = 350 (M+1). 1H NMR (DMSO-d6) 2.1 (s, -CH3), 2.9
3.1.4.2. 5-(4-Hydroxy-6-methyl-2-oxo-2H-pyran-3-yl)-7-(3-methoxyphe-
nyl)-2,3,6,7-tetrahydro-1,4-thiazepine (3). C18H19NO4S calculated m. w.:
345.42. ES-MS: m/z = 346 (M+1).
Pharmazie 66 (2011)
101

ORIGINAL ARTICLES
Table 2: Starting thiazepine ring conformations Cnf and calculated differences of heats of formation ꢀE. Phenyl and pyrane rings
are not displayed for clarity of depiction
Eb [kJ/mol]
Eb [kJ/mol]
Cnf. a S enantiomer
R enantiomer
C1
0.00
13.03
C2
T1
T2
B1
B2
0.00
2.42
13.12
11.93
2.41
11.93
9.77
8.17
2.41c
9.80
a
b
c
Conformation: C chair-like, T twisted, B boat-like. Relative heat of formation related to the minimum found for C2-R: -245.104 kJ/mol. Converted to T1-R during the AM1 calculation
3.1.4.3. 5-(4-Hydroxy-6-methyl-2-oxo-2H-pyran-3-yl)-7-(3,4,5-trimeth-
oxyphenyl)-2,3,6,7-tetrahydro-1,4-thiazepine (4). C20H23NO6S cal-
culated m. w.: 405.47. ES-MS: m/z = 406 (M+1).
3.1.4.6. 5-(4-Hydroxy-6-methyl-2-oxo-2H-pyran-3-yl)-7-(4-chlorophen-
yl)-2,3,6,7-tetrahydro-1,4-thiazepine (7). C17H16ClNO3S calculated m.
w.: 349.84. ES-MS: m/z = 350 (M+1).
3.1.4.4. 5-(4-Hydroxy-6-methyl-2-oxo-2H-pyran-3-yl)-7-(3-hydroxyphe-
nyl)-2,3,6,7-tetrahydro-1,4-thiazepine (5). C17H17NO4S calculated m. w.:
331.39. ES-MS: m/z = 332 (M+1).
3.1.4.7. 5-(4-Hydroxy-6-methyl-2-oxo-2H-pyran-3-yl)-7-(2-nitrophenyl)-
2,3,6,7-tetrahydro-1,4-thiazepine (9). C17H16N2O5S calculated m. w.:
360.39. ES-MS: m/z = 361 (M+1).
3.1.4.5. 5-(4-Hydroxy-6-methyl-2-oxo-2H-pyran-3-yl)-7-(4-hydroxyphe-
nyl)-2,3,6,7-tetrahydro-1,4-thiazepine (6). C17H17NO4S calculated m. w.:
331.39. ES-MS: m/z = 332 (M+1).
3.1.4.8. 5-(4-Hydroxy-6-methyl-2-oxo-2H-pyran-3-yl)-7-(2, 4-dimethox-
yphenyl)-2,3,6,7-tetrahydro-1,4-thiazepine (11). C19H21NO5S calculated
m. w.: 375.44. ES-MS: m/z = 376 (M+1).
102
Pharmazie 66 (2011)

ORIGINAL ARTICLES
3.1.4.9. 5-(4-Hydroxy-6-methyl-2-oxo-2H-pyran-3-yl)-7-(4-isopropylph-
enyl)-2,3,6,7-tetrahydro-1,4-thiazepine (12). C20H23NO3S calculated m.
w.: 357.47. ES-MS: m/z = 358 (M+1).
zolo[4,3-b]pyridazine: a functionally selective gamma-aminobutyric
acid(A) (GABA(A)) alpha2/alpha3-subtype selective agonist that
exhibits potent anxiolytic activity but is not sedating in animal models. J
Med Chem 48: 7089–7092.
Clark T, Alex A, Beck B, Chandrasekhar J, Gedeck P, Horn A, Hutter M,
Martin B, Rauhut G, Sauer W, Schindler T, Steinke T (1999) VAMP (ver-
sion 7.5a). The Oxford Molecular Group The Medawar Centre, Oxford
Science Park, Oxford OX4 4GA, UK.
Clayton T, Chen JL, Ernst M, Richter L, Cromer BA, Morton CJ, Ng H,
Kaczorowski CC, Helmstetter FJ, Furtmuller R, Ecker G, Parker MW,
Sieghart W, Cook JM (2007) An updated unified pharmacophore model of
the benzodiazepine binding site on gamma-aminobutyric acid(a) recep-
tors: correlation with comparative models. Curr Med Chem 14: 2755–
2775.
DaSettimo F, Taliani S, TrincavelliML, Montali M, Martini C (2007)GABA
A/Bz receptor subtypes as targets for selective drugs. Curr Med Chem 14:
2680–2701.
Dawson GR, Collinson N, Atack JR (2005) Development of subtype selec-
tive GABAA modulators. CNS Spectr 10: 21–27.
Drewe J, Kasibhatla S, Tseng B, Shelton E, Sperandio D, Yee
RM, Litvak J, Sendzik M, Spencer JR, Cai SX (2007) Discovery
of 5-(4-hydroxy-6-methyl-2-oxo-2H-pyran-3-yl)-7-phenyl-(E)-2,3,6,7-
tetrahydro-1,4-th iazepines as a new series of apoptosis inducers using
a cell- and caspase-based HTS assay. Bioorg Med Chem Lett 17:
4987–4990.
Granovsky AA (2007) PC GAMESS (version 7.1). Moscow.
Haggarty SJ, Mayer TU, Miyamoto DT, Fathi R, King RW, Mitchison TJ,
Schreiber SL (2000) Dissecting cellular processes using small molecules:
identification of colchicine-like, taxol-like and other small molecules that
perturb mitosis. Chem Biol 7: 275–286.
Hanson SM, Morlock EV, Satyshur KA, Czajkowski C (2008) Structural
requirements for eszopiclone and zolpidem binding to the gamma-
aminobutyric acid type-A (GABAA) receptor are different. J Med Chem
51: 7243–7252.
Hao M-H (2006) Theoretical calculation of hydrogen-bonding strength for
drug molecules. J Chem Theory Comput 2: 863–872.
3.2. Biology
For binding studies, membranes of transiently transfected HEK293 tsa cells
were used. Cells were transfected with GABAA subunits ␣1␤2␥2, ␣2␤3␥2,
␣3␤3␥2, and ␣5␤3␥2 respectively by calcium phosphate DNA precipitation.
24 h after transfection cells were scraped from the plates in PBS and the
membrane fractions prepared after different steps of separation.
Radioactivebinding assayswith 3H-flumazenilwere performed using Multi-
Screen glass fiber filterplates. 10-100 ␮g protein per well was incubated with
1 nM 3H-flumazenil in 50 mM Tris, 200 mM NaCl, pH 7.1 for 60 min at 4 ◦C.
Nonspecific binding was determined in the presence of 10 ␮M diazepam.
Compounds were solubilized in 100 % DMSO and tested in 1 % DMSO at
a range of 6 to 10 different concentrations. Assays were terminated by aspi-
ration using the MultiScreen vacuum manifold and washing with 200 ␮l
assay buffer. After removing the underdrain from the plates, plates were
dryed at 50 ◦C for 2 h. 50 ␮l liquid scintillator was applied to each well and
radioactivity was counted in a Wallac Microplate Beta counter.
3.3. Computational methods
3.3.1. Molecular modelling and conformational analysis
Program Sybyl (SYBYL 2006) was used to sketch molecular structures. For
the conformational analysis of compound 1, 12 starting conformations were
built including two chair-like, twisted, and boat-like conformers for each of
the two enantiomers. Initial models were mechanically minimized applying
the Tripos Force Field. Subsequently, geometries were optimized by means
of the semiempirical procedure AM1 (Clark et al. 1999). Calculated heats
of formation are listed in Table 2. They were directly used to compare the
conformers. Resulting geometries were used for alignments and as input for
DFT calculations.
3.3.2. DFT calculations
The method described by Hao (2006) was applied with the exception that
none of the atoms were fixed. The calculation was performed with the PC
GAMESS program (Granovsky 2007) using the B3LYP functional density
approach with a 6-31++G(d,p) basis set. To correct results for the basis
set superposition error (BSSE), the counterpoise procedure was used, i.e.,
the basis sets of both molecules were included into the calculation of the
energetic contributions of the single molecules. Strength of the hydrogen
bond was calculated as difference between the energy of the energetically
optimized two-molecule system and the (single point) energies of the two
single molecules.
Korpi ER, Mattila MJ, Wisden W, Lüddens H (1997) GABA(A)-receptor
subtypes: clinical efficacy and selectivity of benzodiazepine site ligands.
Ann Med 29: 275–282.
Krepps MK, Parkin S, Atwood DA (2001) Hydrogen Bonding with Sulfur.
Cryst Growth Des 1: 291–297.
Mehta AK, Ticku MK (1999) An update on GABAA receptors. Brain Res
Rev 29: 196–217.
Möhler H, Rudolph U (2004) Selective GABAA circuits for novel CNS
drugs. Drug Disc Today 1: 117–123.
Ohta S, Inujima Y, Abe M, Uosaki Y, Sato S, Miki I (2001) Inhi-
Acknowledgements: The presented investigations were funded by EFRE
grants from the EU and by grants from the Free State of Saxony (project no.
10843).
bition of P-selectin specific cell adhesion by
a low molecular
weight, non-carbohydrate compound, KF38789. Inflamm Res 50: 544–
551.
Qamar Y, Siddiq M (1988) Synthesis of 2-Aryl-7-methylpyrano<4,3-
b>pyran-4(H), 5(H)-diones. Indian Journal of Chemistry 27: 373.
Rachedi Y, Hamdi M, Speziale V (1989) Synthesis of 4-hydroxy 6-
methyl 3-ß-arylpropionyl 2-pyrones by selective catalytic hydrogenation
of 3-cinnamoyl 4-hydroxy 6-methyl 2-pyrones. Synth Commun 19:
3437–3442.
Rudolph U, Crestani F, Benke D, Brünig I, Benson JA, Fritschy JM, Martin
JR, Bluethmann H, Möhler H (1999) Benzodiazepine actions mediated by
specific gamma-aminobutyric acidA receptor subtypes. Nature (London)
401: 796–800.
References
Adman E, Watenpaugh KD, Jensen LH (1975) NH—S hydrogen bonds
in Peptococcus aerogenes ferredoxin, Clostridium pasteurianum rubre-
doxin, and Chromatium high potential iron protein. Proc Natl Acad Sci
U S A 72: 4854–4858.
Albaugh PA, Marshall L, Gregory J, White G, Hutchison A, Ross
PC, Gallagher DW, Tallman JF, Crago M, Cassella JV (2002)
Synthesis and biological evaluation of 7,8,9,10-tetrahydroimidazo[1,2-
c]pyrido[3,4-e]pyrimdin-5(6H)-ones as functionally selective ligands of
the benzodiazepine receptor site on the GABA(A) receptor. J Med Chem
45: 5043–5051.
Atack JR (2005) The benzodiazepine binding site of GABA(A) receptors
as a target for the development of novel anxiolytics. Expert Opin Investig
Drugs 14: 601–618.
Basile AS, Lippa AS, Skolnick P (2004) Anxioselective anxiolytics: can
less be more? Eur J Pharmacol 500: 441–451.
Bonnert TP, Mckernan RM, Farrar S, le Bourdelles B, Heavens RP, Smith
DW, Hewson L, Rigby MR, Sirinathsinghji DJ, Brown N, Wafford KA,
Whiting PJ (1999) Theta, a novel gamma-aminobutyric acid type A recep-
tor subunit. Proc Natl Acad Sci U S A 96: 9891–9896.
Carling RW, Madin A, Guiblin A, Russell MG, Moore KW, Mitchinson
A, Sohal B, Pike A, Cook SM, Ragan IC, Mckernan RM, Quirk K,
Ferris P, Marshall G, Thompson SA, Wafford KA, Dawson GR, Atack
JR, Harrison T, Castro JL, Street LJ (2005) 7-(1,1-Dimethylethyl)-6-
(2-ethyl-2H-1,2,4-triazol-3-ylmethoxy)-3-(2-fluoro phenyl)-1,2,4-tria-
SabinJR(1971)Hydrogenbondsinvolvingsulfur. I. Hydrogensulfidedimer.
J Am Chem Soc 93: 3613–3620.
Savic MM, Huang S, Furtmuller R, Clayton T, Huck S, Obradovic DI,
Ugresic ND, Sieghart W, Bokonjic DR, Cook JM (2007) Are GABA(A)
receptors containing alpha5 subunits contributing to the sedative prop-
erties of benzodiazepine site agonists? Neuropsychopharmacology 33:
332–339.
Scott-Stevens P, Atack JR, Sohal B, Worboys P (2005) Rodent pharmacoki-
netics and receptor occupancy of the GABAA receptor subtype selective
benzodiazepine site ligand L-838417. Biopharm Drug Dispos 26:
13–20.
Sieghart W (2006) GABAA receptors as targets for different classes of drugs.
Drugs Future 31: 685–694.
Sucheta K, Prashant A, Rama Rao N (1995) Synthesis of some new 1,5-
benzothiazepine derivatives. Indian J Chem 34B: 893–894.
SYBYL (2006) (version 7). Tripos Inc. 1699 S. Hanley Road, St. Louis, MO
63144-2913, USA.
Pharmazie 66 (2011)
103

ORIGINAL ARTICLES
Takeuchi N, Nakagawa H, Tobinaga S (1980) Intra- and intermolecular
dimethylether complexes with nitric acid. Chem Phys Lett 391:
condensation reactions of 8-phenyl-7-octene-2,4,6-trione and 8-phenyl-
2,4,6-octanetrione (studies on the ß-carbonyl compounds connected with
the ß-polyketides. IV). Chem Pharm Bull (Tokyo) 28: 3002–3006.
Wierzejewska M (2000) Infrared matrix isolation studies of complexes
formed between dimethylsulfide, dimethyldisulfide and nitrous acid. J
Mol Struct 520: 199–214.
143–147.
Wierzejewska M, Saldyka M (2006) Theoretical study of hydrogen bonded
complexes of dimethyl disulfide or dimethyl peroxide with nitric acid. J
Mol Biol 786: 33–38.
Zhang W, Koehler KF, Zhang P, Cook JM (1995) Development of a com-
prehensive pharmacophore model for the benzodiazepine receptor. Drug
Des Discov 12: 193–248.
Wierzejewska M, Saldyka
M (2004) Are hydrogen bonds to sul-
fur and oxygen different? Theoretical study of dimethylsulfide and
104
Pharmazie 66 (2011)