3-Alkyl- and 3-amido-isothiazoloquinolin-4-ones as ligands for the benzodiazepine site of GABAA receptors

Article abstract of DOI:10.1016/j.bioorg.2011.10.001

Based on a pharmacophore model of the benzodiazepine binding site of the GABA_A receptors, developed with synthetic flavones and potent 3-carbonylquinolin-4-ones, 3-alkyl- and 3-amido-6-methylisothiazoloquinolin-4- ones were designed, prepared and assayed. The suggestion that the interaction between the hydrogen bond donor site H1 with the 3-carbonyl oxygen in 3-carbonylquinolin-4-ones can be replaced by an interaction between H1 and N-2 in the isothiazoloquinolin-4-ones, was confirmed. As with the 3-carbonylquinolin-4-ones, the length of the chain in position 3 is critical for an efficient interaction with the lipophilic pockets of the pharmacophore model. The most potent 3-alkyl derivative, 3-pentyl-6-methylisothiazoloquinolin- 4-one, has an affinity (K_i value) for the benzodiazepine binding site of the GABA_A receptors of 13 nM. However, by replacing the 3-pentyl with a 3-butyramido group an even more potent compound was obtained, with a K_i value of 2.8 nM, indicating that the amide function facilitates additional interactions with the binding site.

Full text of DOI:10.1016/j.bioorg.2011.10.001

Bioorganic Chemistry 40 (2012) 125–130
Contents lists available at SciVerse ScienceDirect
Bioorganic Chemistry
journal homepage: www.elsevier.com/locate/bioorg
3-Alkyl- and 3-amido-isothiazoloquinolin-4-ones as ligands
for the benzodiazepine site of GABAA receptors
Jakob Nilsson a, Elsebet Østergaard Nielsen b, Tommy Liljefors c, Mogens Nielsen c, Olov Sterner a,
⇑
a Division of Organic Chemistry, Lund University, PO Box 124, SE-221 00 Lund, Sweden
b NeuroSearch A/S, DK-2750 Ballerup, Denmark
c University of Copenhagen, Faculty of Pharmaceutical Sciences, 2 Universitetsparken, DK-2100 Copenhagen, Denmark
a r t i c l e i n f o
a b s t r a c t
Article history:
Received 13 September 2011
Available online 18 October 2011
Based on a pharmacophore model of the benzodiazepine binding site of the GABAA receptors, developed
with synthetic flavones and potent 3-carbonylquinolin-4-ones, 3-alkyl- and 3-amido-6-methylisothiazol-
oquinolin-4-ones were designed, prepared and assayed. The suggestion that the interaction between the
hydrogen bond donor site H1 with the 3-carbonyl oxygen in 3-carbonylquinolin-4-ones can be replaced
by an interaction between H1 and N-2 in the isothiazoloquinolin-4-ones, was confirmed. As with the
3-carbonylquinolin-4-ones, the length of the chain in position 3 is critical for an efficient interaction with
the lipophilic pockets of the pharmacophore model. The most potent 3-alkyl derivative, 3-pentyl-6-meth-
ylisothiazoloquinolin-4-one, has an affinity (Ki value) for the benzodiazepine binding site of the GABAA
receptors of 13 nM. However, by replacing the 3-pentyl with a 3-butyramido group an even more potent
compound was obtained, with a Ki value of 2.8 nM, indicating that the amide function facilitates addi-
tional interactions with the binding site.
Keywords:
3-Alkyl-6-methylisothiazolo[5,4-
b]quinolin-4(9H)-ones
3-Amido-6-methylisothiazolo[5,4-
b]quinolin-4(9H)-ones
Benzodiazepine binding site
GABAA receptors
Pharmacophore model
Ó 2011 Elsevier Inc. All rights reserved.
1. Introduction
The affinity (expressed as the Ki value) of 1, 2 as well as 3 for the ben-
zodiazepine binding site is approximately 1 nM [5,6].
The major inhibitory neurotransmitter in the central nervous
system is -aminobutyric acid (GABA) [1]. Ionotropic receptors
Fig. 2 shows the present pharmacophore model (left), and com-
pound 1 positioned in the pharmacophore model (right). The 3-
carbonyl oxygen of 1 is believed to interact with the hydrogen
bond donating site H1 in the pharmacophore model, and in a
search for new scaffolds for binding site we believed that this
interaction can be replaced with that between H1 and N-2 in suit-
ably substituted isothiazoloquinolin-4-ones. As can be seen in the
pharmacophore model (Fig. 2, left) there is little room for substit-
uents, essentially only the lipophilic pocket towards L2 and the
c
for GABA are ligand gated ion channels that on activation by GABA
mediate fast neurotransmission by allowing a flow of chloride ions
into the neuron, causing a hyperpolarization of the membrane and
inhibiting further neuronal activity. Allosteric modulatory sites on
GABAA receptors exist for various ligands including benzodiaze-
pine, picrotoxin, loreclezole, ethanol, barbiturate and zinc cations,
of which the benzodiazepine site has attracted most attention.
Full agonists acting at the benzodiazepine site have long been
used as anxiolytics, although their applicability is limited due to ad-
verse effects such as sedation, cognitive impairment and ataxia. An
earlier pharmacophore model of the benzodiazepine binding site
[2] has been developed and refined based on SAR studies of synthetic
flavone derivatives [3,4]. The model was recently applied for the
identification and optimization of novel 3-carbonylquinolin-4-ones
[5,6], azaflavones [7] and triazoloquinazolinediones [8] as ligands at
the GABAA receptors. Fig. 1 shows the structures of three 3-carbon-
ylquinolin-4-one derivatives that bind strongly to the benzodiaze-
pine binding site of the GABAA receptors, 3-valeryl-6-
methylquinolin-4-one 1, 3-propyloxycarbonyl-6-methylquinolin-
4-one 2 and 3-(N-propylcarbamoyl)-6-methylquinolin-4-one 3.
interface region (an area between an
a and a c subunit in the pen-
taheteromeric receptor GABAA). In this study we have focussed on
the lipophilic pocket and kept the substituent in position 6 con-
stant. Consequently, compounds 13, 14 and 15 were designed, syn-
thesized, and assayed. As the presence of a carbamoyl function at
position 3 in the quinolin-4-ones (e.g. 3, Fig. 1) in some cases have
been shown to be advantageous, we decided to include compound
18 and 19 as well (having an amido group instead of a carbamoyl
group in position 3).
2. Material and methods
2.1. General procedures for synthesis
⇑
Reagents and solvents (except THF) were used from commercial
sources without purification. THF was distilled from sodium/
Corresponding author. Fax: +46 462228209.
E-mail address: Olov.Sterner@organic.lu.se (O. Sterner).
0045-2068/$ - see front matter Ó 2011 Elsevier Inc. All rights reserved.
doi:10.1016/j.bioorg.2011.10.001

126
J. Nilsson et al. / Bioorganic Chemistry 40 (2012) 125–130
Fig. 1. Potent 3-carbonyl-6-ethylisothiazoloquinolinones representing the starting point for this study.
Fig. 2. Left: The pharmacophore model used for the design of the new structures. H1 and H2 are hydrogen bond donor sites while A2 and A3 are hydrogen bond acceptor
sites. L1, L2 and L3 represent lipophilic pockets and S1–S5 denotes regions of steric repulsive ligand-receptor interactions (receptor essential volume). The interface region is a
partly lipophilic region that has been suggested to represent the interface between
positioned in the model.
a- and c-subunits in GABAA receptors. Right: 3-valeryl-6-ethylisothiazoloquinolinone (1)
benzophenone prior to use. 1H and 13C NMR were recorded at room
temperature unless otherwise specified with a Bruker DR400 spec-
trometer. The spectra were recorded in CDCl3, DMSO-d6, and C6D6,
and the solvent signals (7.27 and 77.0, 2.50 and 39.5 or 7.18 and
128.1 ppm, respectively) were used as reference. Analytical thin
layer chromatography (TLC) was performed on Kiselgel 60 F254
plates (Merck). Column chromatography was performed on SiO2
(Matrex LC-gel: 60A, 35-70 MY, Grace). Melting points (uncor-
rected) were determined with a Reichert microscope. EI mass spec-
tra were recorded at 70 eV with a Jeol SX102 spectrometer and ESI
spectra were recorded with Micromass Q-TOF Micro.
product was recrystallized from ethanol to give 6 as white needle-
shaped crystals (2.17 g, 83%). mp: 114 °C. 1H NMR (400 MHz,
CDCl3 + 5% MeOD-d4) d 7.95 (1H, s), 7.57 (2H, bs), 2.70 (3H, s),
2.45 (3H, s); 13C NMR (100 MHz, CDCl3 + 5% MeOD-d4) d 183.5,
162.3, 146.3, 138.0, 137.1, 129.8, 124.7, 119.1, 21.3, 14.2; HRMS
(ESI): for C10H10NOS2 Calcd: 224.0204; [M+H]; found: 224.0201.
2.1.3. Methyl {2-[(2Z)-3-hydroxypent-2-enoyl]-4-methylphenyl}
dithiocarbamate (7)
A solution of 1.6 M n-BuLi (0.78 mL, 1.25 mmol) in hexane was
added to a solution of diisopropylamine (0.18 mL, 1.3 mmol) in
5 mL of THF under N2 atmosphere at À78 °C. The solution was
heated to 0 °C and stirred for 5 min and then once again cooled
to À78 °C. To the resultant LDA solution was added a solution of
butanone (0.112 mL, 1.25 mmol) in 2 mL of THF and the mixture
was stirred for 1 h. A solution of 6 in 3 mL of THF was slowly added
and the mixture was slowly heated to À40 °C over a period of 3 h,
while monitored by TLC. The reaction was poured onto 3 mL of an
aqueous solution of HCl (1 M) and the mixture was concentrated to
half its volume. The residue was extracted once with 100 mL of
EtOAc and the organic layer was washed with brine, dried over
MgSO4 and concentrated under reduced pressure. The crude prod-
uct was purified by chromatography on silica gel column. Elution
with n-heptane/toluene/acetone (50:50:1) as eluent yielded 7
(61%) as a yellow solid (mp 25 °C). 1H NMR (400 MHz, CDCl3) d
11.70 (1H, s), 8.63 (1H, d, J = 8.4 Hz), 7.49 (1H, J = 1.7 Hz), 7.33
(1H, dd, J = 8.4 and 1.7 Hz), 6.09 (1H, s), 2.67 (3H, s), 2.43 (2H, q,
J = 7.6 Hz), 2.38 (3H, s), 1.24 (3H, t, J = 7.6 Hz); 13C NMR
(100 MHz, CDCl3) d 198.4, 192.5, 191.7, 137.1, 135.2, 133.4,
129.5, 126.3, 124.4, 97.4, 30.6, 21.2, 18.7, 10.5; HRMS (ESI): for
2.1.1. 5-Methyl-2-{[(methylsulfanyl)carbonothioyl]amino}benzoic
acid (5)
To a solution of 2-amino-5-methylbenzoic acid (4, 2.52 g,
16.7 mmol) and carbon disulfide (2.01 mL, 33.9 mmol) in 45 mL
of dry 1,4-dioxane was added NEt3 (5.58 mL, 40.0 mmol) and the
mixture were stirred under N2 atmosphere at 5 °C for 18 h. Iodo-
methane (1.14 mL, 18.4 mmol) was added dropwise and the mix-
ture was stirred at 5 °C for 1 h. The reaction was poured into
25 mL of an aqueous solution of HCl (1 M) and the mixture was
concentrated to half its volume under reduced pressure and ex-
tracted three times with EtOAc (75 mL each time). The combined
organic layers were dried over MgSO4 and concentrated under re-
duced pressure. The residue was recrystallized from chloroform to
give 5 as a yellow solid (3.25 g, 81%). mp: 199 °C. 1H NMR
(400 MHz, MeOD-d4) d 8.82 (1H, d, J = 8.2 Hz), 7.90 (1H, s), 7.38
(1H, d, J = 8.2 Hz), 2.63 (3H, s), 2.35 (3H, s); 13C NMR (100 MHz,
MeOD-d4) d 189.2, 161.4, 130.9, 126.5, 125.4, 123.2, 114.2, 110.7,
11.3, 8.9; HRMS (ESI): for C10H12NO2S2 Calcd: 242.0309; [M+H];
found: 242.0316.
C14H18NO2S2 Calcd: 296.0779; [M+H]; found: 296.0763.
2.1.2. 6-Methyl-2-(methylsulfanyl)-4H-3,1-benzothiazin-4-one (6)
Compound 5 (2.85 g, 11.8 mmol) was dissolved in 50 mL of ace-
tic anhydride and heated at reflux for 1 h. The mixture was cooled
to room temperature and the precipitate was filtered off. The crude
2.1.4. Methyl {2-[(2Z)-3-hydroxyhept-2-enoyl]-4-methylphenyl}
dithiocarbamate (8)
Methyl
{2-[(2Z)-3-hydroxyhept-2-enoyl]-4-methylphenyl}
dithiocarbamate (8) was prepared and purified according to the

J. Nilsson et al. / Bioorganic Chemistry 40 (2012) 125–130
127
description for 7, starting from 2-hexanone. The reaction yielded 8
(75%) as a yellow solid (mp 58 °C). 1H NMR (400 MHz, CDCl3) d
15.64 (1H, s), 11.81 (1H, s), 8.66 (1H, d, J = 8.4 Hz), 7.48 (1H,
J = 1.7 Hz), 7.31 (1H, dd, J = 8.4 and 1.7 Hz), 6.09 (1H, s), 2.67 (3H,
s), 2.37 (5H, m), 1.66 (2H, m), 1.41 (2H, hex., J = 7.6 Hz), 0.96 (3H,
t, J = 7.3 Hz); 13C NMR (100 MHz, CDCl3) d 198.2, 192.6, 190.6,
137.2, 135.0, 133.4, 129.5, 126.0, 124.1, 98.1, 37.0, 28.5, 22.5,
21.1, 18.6, 13.9; HRMS (ESI): for C16H22NO2S2 Calcd: 324.1092;
[M+H]; found: 324.1101.
HRMS (ESI): for C16H20NO2S Calcd: 290.1209; [M+H]; found:
290.1209.
2.1.9. 3-Ethyl-6-methylisothiazolo[5,4-b]quinolin-4(9H)-one (13)
To a solution of 10 (64 mg, 0.259 mmol) in 25 mL of methanol
was added
a
solution of hydroxylamine-O-sulfonic acid
(102.6 mg, 0.907 mol) and lithium hydroxide (38.1 mg,
0.907 mmol) in 3 mL of methanol and the mixture was stirred at
room temperature for 30 h. The reaction mixture was concentrated
under reduced pressure and applied to flash chromatography. Elu-
tion with heptane/EtOAc (3:1) yielded 13 (63%) as white crystals
(mp 332 °C). 1H NMR (400 MHz, DMSO-d6) d 12.70 (1H, s), 8.03
(1H, d, J = 2.0 Hz), 7.58 (1H, dd, J = 8.4 and 2.0 Hz), 7.43 (1H, d,
J = 8.4 Hz), 3.17 (2H, q, J = 7.4 Hz), 2.43 (3H, s), 1.26 (3H, t,
J = 7.4 Hz); 13C NMR (100 MHz, DMSO-d6) d 173.4, 170.3, 165.8,
138.1, 134.4, 132.3, 125.3, 123.3, 118.0, 117.5, 27.1, 20.6, 12.0;
HRMS (ESI): for C13H13N2OS Calcd: 245.0749; [M+H]; found:
245.0747.
2.1.5. Methyl {2-[(2Z)-3-hydroxyoct-2-enoyl]-4-methylphenyl}
dithiocarbamate (9)
Methyl {2-[(2Z)-3-hydroxyoct-2-enoyl]-4-methylphenyl}dithio
carbamate (9) was prepared and purified according to the descrip-
tion for 7, starting from 2-heptanone. The reaction yielded 9 (51%)
as a yellow solid (mp 63 °C). 1H NMR (400 MHz, CDCl3) d 15.66 (1H,
s), 11.79 (1H, s), 8.67 (1H, d, J = 8.4 Hz), 7.49 (1H, d, J = 1.6 Hz), 7.34
(1H, dd, J = 8.4 and 1.6 Hz), 6.10 (1H, s), 2.68 (3H, s), 2.39 (5H, m),
1.7 (2H, m), 1.37 (4H, m), 0.93 (3H, m); 13C NMR (100 MHz, CDCl3)
d 198.3, 192.7, 190.7, 137.3, 135.1, 133.4, 129.6, 126.2, 124.3, 98.2,
37.4, 31.6, 26.2, 22.6, 21.2, 18.7, 14.1; HRMS (ESI): for C17H24NO2S2
Calcd: 338.1243; [M+H]; found: 338.1267.
2.1.10. 3-Butyl-6-methylisothiazolo[5,4-b]quinolin-4(9H)-one (14)
3-Butyl-6-methylisothiazolo[5,4-b]quinolin-4(9H)-one
(14)
was prepared and purified according to the description for 13,
starting from 11. The reaction yielded 14 (82%) as white crystals
(mp 243 °C). 1H NMR (400 MHz, DMSO-d6) d 12.72 (1H, s), 8.02
(1H, s), 7.56 (1H, d, J = 8.4 Hz), 7.41 (1H, d, J = 8.4 Hz), 3.15 (2H, t,
J = 7.7 Hz), 2.42 (3H, s), 1.69 (2H, pent, J = 7.5 Hz), 1.36 (2H, hex,
J = 7.4 Hz), 0.91 (3H, t, J = 7.4 Hz); 13C NMR (100 MHz, DMSO-d6)
d 173.4, 169.3, 165.7, 138.0, 134.4, 132.2, 125.3, 123.5, 118.1,
117.4, 33.1, 29.6, 21.9, 20.6, 13.8; HRMS (ESI): for C15H17N2OS
Calcd: 273.1062; [M+H]; found: 273.1077.
2.1.6. 1-(4-Hydroxy-6-methyl-2-thioxo-1,2-dihydroquinolin-3-yl)
propan-1-one (10)
To keto-enol 7 (91.3 g, 0.309 mmol) was added 5 mL of a 0.5 M
solution of sodium methoxide in methanol and the mixture was
stirred at 0 °C for 3 h. A 1.0 M solution of hydrochloric acid
(3 mL) was poured onto the reaction and the mixture was concen-
trated to less than half its volume under reduced pressure and ex-
tracted three times with EtOAc. The combined organic layers were
dried over MgSO4, concentrated under reduced pressure and pre-
cipitated from MeOH to give 10 as a yellow solid (1.70 g, 97%).
mp: 238 °C; 1H NMR (400 MHz, DMSO-d6) d 13.11 (1H, s), 12.27
(1H, s), 7.91 (1H, s), 7.52 (2H, s), 2.93 (2H, q, J = 7.1 Hz), 2.38 (3H,
s), 1.09 (3H, s); 13C NMR (100 MHz, CDCl3) d 204.2, 176.8, 156.7,
138.2, 133.9, 133.1, 123.7, 122.7, 117.0, 116.3, 36.4, 20.8, 7.9;
HRMS (ESI): for C13H14NO2S Calcd: 248.0745; [M+H]; found:
248.0739.
2.1.11. 6-Methyl-3-pentylisothiazolo[5,4-b]quinolin-4(9H)-one (15)
6-Methyl-3-pentylisothiazolo[5,4-b]quinolin-4(9H)-one
(15)
was prepared and purified according to the description for 13,
starting from 12. The reaction yielded 15 (93%) as white crystals
(mp 245 °C). 1H NMR (400 MHz, DMSO-d6) d 12.71 (1H, s), 8.03
(1H, s), 7.57 (1H, dd, J = 8.4 and 2.0 Hz), 7.42 (1H, d, J = 8.4 Hz),
3.14 (2H, m), 2.42 (3H, s), 1.71 (2H, m), 1.33 (4H, m), 0.87 (3H,
m); 13C NMR (100 MHz, DMSO-d6) d 173.4, 169.4, 165.8, 138.0,
134.4, 132.2, 125.3, 123.5, 118.2, 117.4, 33.4, 31.1, 27.2, 21.9,
20.6, 13.9; HRMS (ESI): for C16H19N2OS Calcd: 287.1213; [M+H];
found: 287.1210.
2.1.7. 1-(4-Hydroxy-6-methyl-2-thioxo-1,2-dihydroquinolin-3-yl)
pentan-1-one (11)
1-(4-Hydroxy-6-methyl-2-thioxo-1,2-dihydroquinolin-3-
yl)pentan-1-one (11) was prepared and purified according to the
description for 10, starting from 8. The reaction yielded 11 (97%)
as a yellow solid (mp 170 °C). 1H NMR (400 MHz, CDCl3) d 17.13
(1H, s), 11.18 (1H, s), 7.96 (1H, d, J = 1.7 Hz), 7.50 (1H, dd, J = 8.4
and 1.7 Hz), 7.27 (1H, d, J = 8.4 Hz), 3.76 (2H, t, J = 7.3 Hz), 2.45
(3H, s), 1.77 (2H, pent, J = 7.4 Hz), 1.47 (2H, hex, J = 7.4 Hz), 0.98
(3H, t, J = 7.4 Hz); 13C NMR (100 MHz, CDCl3) d 211.5, 180.6,
173.0, 138.1, 136.9, 134.6, 125.0, 117.6, 115.9, 115.4, 44.5, 27.2,
22.6, 21.4, 14.3; HRMS (ESI): for C15H18NO2S Calcd: 276.1058;
[M+H]; found: 276.1072.
2.1.12. Methyl [2-(cyanoacetyl)-4-methylphenyl]dithiocarbamate
(16)
A solution of 1.6 M n-BuLi (5.4 mL, 13.4 mmol) in hexane was
added to a solution of diisopropylamine (1.95 mL, 14.0 mmol) in
20 mL of THF under N2 atmosphere at À78 °C. The solution was
heated to 0 °C and stirred for 5 min and then once again cooled
to À78 °C. To the resultant LDA solution was added acetonitrile
(0.28 mL, 13.4 mmol) and the mixture was stirred for 1 h at
À20 °C. A solution of 6 (1.20 g, 5.4 mmol) in 8 mL of THF was
slowly added at À78 °C and the mixture was stirred for 1 h. The
reaction was poured onto 40 mL of an aqueous solution of HCl
(1 M) and the mixture was concentrated to half its volume. The
residue was extracted with EtOAc and the organic layer was
washed with brine, dried over MgSO4, concentrated and crystal-
lized from ethyl alcohol to give 16 as a yellow solid (1.41 g, 99%).
mp: 142 °C; 1H NMR (400 MHz, DMSO-d6) d 1H NMR (400 MHz,
CDCl3) d 11.89 (1H, s), 9.00 (1H, d, J = 8.3 Hz), 7.50 (1H, s), 7.48
(1H, d, J = 8.3 Hz), 4.17 (2H, s), 2.41 (3H, s); 13C NMR (100 MHz,
CDCl3) d 199.3, 191.0, 139.2, 136.7, 135.0, 130.6, 123.5, 122.1,
2.1.8. 1-(4-Hydroxy-6-methyl-2-thioxo-1,2-dihydroquinolin-3-yl)
hexan-1-one (12)
1-(4-Hydroxy-6-methyl-2-thioxo-1,2-dihydroquinolin-3-yl)
hexan-1-one (12) was prepared and purified according to the
description for 10, starting from 9. The reaction yielded 12 (96%)
as a yellow solid (mp 150 °C). 1H NMR (400 MHz, CDCl3) d 17.19
(1H, s), 11.39 (1H, s), 7.99 (1H, bs), 7.54 (1H, dd, J = 8.4 and
1.6 Hz), 7.32 (1H, d, J = 8.4 Hz), 3.67 (2H, t, J = 7.3 Hz), 2.45 (3H,
s), 1.79 (2H, pent, J = 7.3 Hz), 1.41 (4H, m), 0.94 (3H, t, J = 7.2 Hz);
13C NMR (100 MHz, CDCl3) d 211.3, 179.7, 173.1, 138.1, 137.0,
134.9, 124.9, 117.7, 115.7, 115.6, 44.7, 31.7, 24.8, 22.8, 21.4, 14.2;
113.5, 31.2, 21.2, 18.7; HRMS (ESI): for
265.0469; [M+H]; found: 265.0483.
C12H13N2OS2 Calcd:

128
J. Nilsson et al. / Bioorganic Chemistry 40 (2012) 125–130
2.1.13. 4-Hydroxy-6-methyl-2-thioxo-1,2-dihydroquinoline-3-
carbonitrile (17)
GF/C glass fiber filters under suction and immediately washed with
5 mL ice-cold buffer. Non-specific binding was determined by add-
4-Hydroxy-6-methyl-2-thioxo-1,2-dihydroquinoline-3-carbo-
nitrile (17) was prepared and purified according to the description
for 10, starting from 16. The reaction yielded 17 (97%) as a yellow
solid (mp 284 °C). 1H NMR (400 MHz, DMSO-d6) d 12.89 (1H, s),
7.85 (1H, s), 7.51 (1H, d, J = 8.3 Hz), 7.44 (1H, d, J = 8.3 Hz), 2.36
(3H, s); 13C NMR (100 MHz, DMSO-d6) d 171.2, 164.9, 138.2,
135.1, 134.7, 123.7, 121.0, 118.1, 116.0, 98.8, 20.8; HRMS (ESI):
for C11H9N2OS Calcd: 217.0436; [M+H]; found: 217.0439.
ing Clonazepam (1 lM final concentration) to separate samples.
Protein was estimated by conventional protein assay method using
Bovine serum albumin as standard. IC50 values were determined by
assaying 4–6 different concentrations of each test substance. Ki val-
ues were calculated according to Ki = IC50/(1 + (L)/KD), (L) is the
concentration (nM) of 3H-flumazenil; KD is binding affinity con-
stant of 3H-flumazenil (1.6 nM).
3. Results
2.1.14. N-(6-Methyl-4-oxo-4,9-dihydroisothiazolo[5,4-b]quinolin-3-
yl)propanamide (18)
3.1. Chemical synthesis
To a solution of 17 (30 mg, 0.139 mmol) in 1 mL of DMF was
added a solution of NaHCO3 (41 mg, 0.43 mmol) and hydroxyl-
amine-O-sulfonic acid (55 mg, 0.43 mmol) in 1 mL of methyl alco-
hol and the mixture was stirred for 2 h, during which a white
precipitate presumed to be the aminothioisoxazole was formed.
This was filtered off and slurried in 2 mL of DMF. Propionyl chlo-
All isothiazolo[5,4-b]quinolin-4(9H)-ones prepared and pre-
sented in this investigation are to our knowledge new compounds.
For all target compounds, an unambiguous structure determina-
tion was performed using COSY, HMQC, HMBC and NOESY NMR
experiments as well as high resolution mass spectrometry.
ride (19.2 lL, 0.221 mmol) and 0.10 mL of pyridine was added
The key intermediate for the preparation of the isothiazolo[5,4-
b]quinolin-4(9H)-one derivatives 13, 14, 15, 18 and 19 is the
4H-benzo[d][1,3]thiazin-4-one 6, which was synthesized by the
addition of carbon disulfide to the anthranilic acid 4 followed by
methylation of the sulfur and thiolactone formation in acetic anhy-
dride (Scheme 1) [9]. The addition of different 2-alkanones to 6
afforded the keto-enols 7–9, which were recyclized to give the cor-
responding thione derivatives 10–12. Treatment of the thiones 10–
12 with hydroxylamine-O-sulfonic acid afforded the 3-alkyl-6-
methyl-isothiazolo[5,4-b]quinolin-4(9H)-ones 13–15 [10] (Scheme
1). The 3-amido-6-methylisothiazolo[5,4-b]quinolin-4(9H)-ones
18 and 19 were prepared from the cyano derivative 16, prepared
by the addition of acetonitile to 6 (Scheme 2) following essentially
the same protocol as for the synthesis of 7, and cyclized to 17. The
amination of the thione functionality of 17 with hydroxylamine-O-
sulfonic acid yielded the aminothioisoxazole, which due to limited
chemical stability was not isolated but acylated directly with pro-
pionyl chloride and butyryl chloride to yield 18 and 19, respec-
tively. Cromatographic and spectroscopic investigations of the
degradation products of the aminothioisoxazole formed from 17
suggested that it gradually polymerized in DMF.
and a clear solution was formed. The mixture was stirred for 5 h,
concentrated and purified by chromatography. Elution with n-hep-
tane/EtOAc (1:1) afforded 18 (19%) as a white solid [mp 300 °C (de-
comp)]. 1H NMR (400 MHz, DMSO-d6) d 11.62 (1H, s), 9.80 (1H, s),
7.86 (1H, d, J = 2 Hz), 7.74 (1H, dd, J = 8.4 and 2 Hz), 7.59 (1H, d,
J = 8.4 Hz), 2.41 (5H, m), 1.07 (3H, t, J = 7.4 Hz); 13C NMR
(100 MHz, DMSO-d6) d 174.2, 171.2, 164.0, 153.0, 138.2, 135.2,
133.3, 124.7, 122.4, 118.0, 110.0, 28.2, 20.8, 9.8; HRMS (FAB+):
for C14H14N3O2S Calcd: 288.0807; [M+H]; found: 288.0832.
2.1.15. N-(6-Methyl-4-oxo-4,9-dihydroisothiazolo[5,4-b]quinolin-3-
yl)butanamide (19)
N-(6-Methyl-4-oxo-4,9-dihydroisothiazolo[5,4-b]quinolin-3-yl)
butanamide (19) was prepared and purified according to the
description for 18, starting from 17, with butyryl chloride
(23.1 lL, 0.221 mmol) instead of propionyl chloride. The reaction
yielded 19 as a white solid [19 mg, 45%, mp 320 °C (decomp)]. 1H
NMR (400 MHz, DMSO-d6) d 12.93 (1H, s), 10.74 (1H, s), 8.00
(1H, d, J = 2.0 Hz), 7.62 (1H, dd, J = 8.4 and 2.0 Hz), 7.47 (1H, d,
J = 8.4 Hz), 2.60 (2H, J = 7.4 Hz), 2.42 (3H, s), 1.65 (2H, hex,
J = 7.4 Hz), 0.95 (3H, t, J = 7.4 Hz); 13C NMR (100 MHz, DMSO-d6)
d 174.2, 170.9, 164.1, 153.3, 138.5, 135.2, 133.2, 124.7, 122.6,
118.1, 109.8, 38.7, 20.7, 18.0, 13.7; HRMS (FAB+): for C15H16N3O2S
Calcd: 302.0963; [M+H]; found: 302.0959.
3.2. Receptor binding
Affinities for the BZD binding site were determined with the
same in vitro radio ligand assay that was used in our previous
investigations, by the displacement of [3H]-flumazenil in rat corti-
cal tissue (see Experimental for details). The results are presented
in Table 1.
2.2. Benzodiazepine receptor binding in vitro
The binding of 3H-flumazenil (87 Ci/mmol) to rat cortical mem-
branes was done following the methods previously described in
detail by Dekermendjian et al. [3]. In brief: Tissue is homogenized
in 20 mL Tris, HCl (30 mM, pH 7.4) using an Ultra-Turrax homoge-
nizer. The suspensions are centrifuged at 27,000g for 15 min fol-
lowed by three centrifugations resuspensions cycles. The washed
pellet is resuspended in 20 mL buffer, incubated at 37 °C for
30 min and then centrifuged for 10 min (27,000g). The pellet is
washed once and the final pellet is resuspended in 30 mL Tris,
HCl buffer (50 mM, pH 7.1) and stored at À20 °C until use. For
binding studies frozen membrane suspensions were thawed and
centrifuged (27,000g, 10 min). The pellet was resuspended into
4. Discussion
The pharmacophore model for ligands of the benzodiazepine
binding site of the GABAA receptors has been used and validated
for the design of several active compound classes [5,6], and again
it is demonstrated that a structure that fits into the model and ful-
fills the requirements discussed below will display affinity for the
benzodiazepine binding site. This is shown in Fig. 3, where 3-pen-
tyl-6-methylisothiazoloquinolin-4-one (15) (left) and 3-butyrami-
do-6-methyl isothiazoloquinolin-4-one (19) (right) are positioned
into the pharmacophore model. The interactions between all
hydrogen bond donator/acceptor sites are satisfied, and no steric
repulsive ligand-receptor interaction is indicated. The small differ-
ence in the Ki value for 15 and 19 noted, approximately five times,
may not completely reflect the difference in affinity of the two
compounds to the binding site as 19 should be considerably more
Tris, citrate buffer (50 mM, pH 7.1) at the concentration 50
tein/0.55 mL assay (1 mg original tissue/0.55 mL assay). Aliquots of
0.5 mL membrane preparation are added to 25
L of 3H-flumazenil
solution (1 nM final concentration) and 25 L containing test sub-
lg pro-
l
l
stance and incubated at an ice-bath (0–4 °C) for 40 min. The incu-
bated samples were added to 5 mL ice-cold buffer (Tris, citrate,
50 mM pH 7.1), the suspension was poured directly onto Whatman

J. Nilsson et al. / Bioorganic Chemistry 40 (2012) 125–130
129
Scheme 1. Conditions: (a) CS2, NEt3, dioxane, 5 °C, 18 h, then MeI, 5 °C, 1 h, yield 81%; (b) Ac2O, reflux, 1 h, yield 83%; (c) LDA, 2-butanone for 7,2-hexanone for 8, or 2-
heptanone for 9, THF, À78 °C, 1 h, then 6, À78 °C to À40 °C, 3 h (yields 61% for 7, 75% for 8, and 51% for 9); (d) NaOMe, MeOH, 0 °C, 3 h, yield 97% for 10, 11 and 12); (e)
H2NOSO3H, LiOH, MeOH, rt, 24 h (yield 63% for 13, 82% for 14, and 93% for 15).
Scheme 2. Conditions: (a) LDA, MeCN, THF, À78 °C, 1 h, then 6, À20 °C, 1 h, yield 99%; (b) NaOMe, MeOH, 0 °C, 3 h, yield 97%; (c) H2NOSO3H, NaHCO3, MeOH, DMF, rt., 2 h,
then CH3CH2COCl for 18 or CH3(CH2)2COCl for 19, pyridine, DMF, 5 h (yield 19% for 18 and 45% for 19.
3-butyl derivative 14 has a slightly higher affinity than the propi-
onamido derivative 18.
Table 1
Ki values of isothiazoloquinolin-4-ones 13, 14, 15, 18 and 19
tested on 3H-flumazenil binding in vitro to rat cortical membranes.
In addition, as has been demonstrated during the development
of both the flavones and the quinolin-4-ones, an efficient filling of
the L2 lipophilic pocket may influence the affinity considerably.
This is confirmed in this investigation, both for the 3-alkyl and
the 3-amido derivatives. Previous investigations (flavones as well
as quinolin-4-ones) have very clearly shown that the space avail-
able at L2 is quite limited and too big substituents in position 3 will
diminish the affinity. For the quinolin-4-ones, nothing bigger than
the 3-substituents in compounds 1, 2 and 3, i.e. a 3-valeryl, a 3-
propyloxycarbonyl or a 3-(N-propylcarbamoyl), will be accepted
by the binding site, independent if it is a longer chain length, a
branched chain, or a cycloalkyl group. We are therefore fairly cer-
tain that the pentyl derivative 15 and the butyroamide derivative
19 are the most potent of these two compound classes.
Compd
Ki (nM)a
13
14
15
18
19
930 120
56
2
13 1.7
121
9
2.8 0.7
a
Each Ki value is the mean of SD of three determinations.
hydrophilic compared to 15. Nevertheless, it is reasonable to as-
sume that 19 is able to bind stronger that 15 due to the extra inter-
action between the amido oxygen and H1 as indicated in Fig. 3
(right), possibly directed by an internal hydrogen bond between
the amido NH and the 4-carbonyl oxygen. However, it cannot be
excluded that this additional interaction actually take place be-
tween the amido oxygen and H2, and it should be noted that the
Another requirement for ligands of the benzodiazepine site ful-
filled by the isothiazoloquinolin-4-ones concerns their three
dimensional form, as previous studies have emphasized that li-

130
J. Nilsson et al. / Bioorganic Chemistry 40 (2012) 125–130
Fig. 3. Left: 3-Pentyl-6-methylisothiazoloquinolinone (15) positioned in the model. Right: 3-Butyramido-6-methylisothiazoloquinolinone (19) positioned in the model.
gands with a strong affinity for this binding site must be able to
adopt a planar or close to planar conformation [2]. There are obvi-
ously no intramolecular interactions possible in the compounds as-
sayed in this investigation that will prevent them from being
essentially planar.
School for Pharmaceutical Sciences at Lund University, Carlsberg
Foundation, Denmark, and the NeuroScience PharmaBiotec Re-
search Center, Denmark, is gratefully acknowledged.
References
[1] W. Sieghart, Adv. Pharmacol. 54 (2006) 231–263.
[2] W. Zhang, K.F. Koehler, P. Zhang, J.M. Cook, Drug Des. Discov. 12 (1995) 193–
248.
5. Conclusion
[3] K. Dekermendjian, P. Kahnberg, M. Witt, O. Sterner, M. Nielsen, T. Liljefors, J.
Med. Chem. 42 (1999) 4343–4350.
The usefulness of the pharmacophore model describing the
binding site for benzodiazepines in the GABAA receptors was again
demonstrated, by the design, synthesis and testing of 3-alkyliso-
thiazolo[5,4-b]quinolin-4(9H)-ones and 3-amidoisothiazolo[5,4-
b]quinolin-4(9H)-ones. The most potent compounds prepared have
high affinity with Ki values in the low nM range. The difference be-
tween the most potent 3-alkylisothiazoloquinolin-4-one (15) and
the most potent 3-amidoisothiazoloquinolin-4-one (19) is sug-
gested to depend on the additional interaction with the H1 site
of the pharmacophore model made possible by the amido oxygen.
[4] P. Kahnberg, E. Lager, C. Rosenberg, J. Schougaard, L. Camet, O. Sterner, E.
Østergaard, M. Nielsen, T. Liljefors, J. Med. Chem. 45 (2002) 4188–4201.
[5] E. Lager, P. Andersson, J. Nilsson, I. Pettersson, E. Østergaard Nielsen, M.
Nielsen, O. Sterner, T. Liljefors, J. Med. Chem. 49 (2006) 2526–2533.
[6] E. Lager, J. Nilsson, E. ØstergaardNielsen, M. Nielsen, T. Liljefors, O. Sterner,
Bioorg. Med. Chem. 16 (2008) 6936–6948.
[7] J. Nilsson, E. Østergaard Nielsen, T. Liljefors, M. Nielsen, O. Sterner, Bioorg. Med.
Chem. Lett. 18 (2008) 5713–5716.
[8] J. Nilsson, E. Østergaard Nielsen, T. Liljefors, M. Nielsen, O. Sterner, Bioorg. Med.
Chem. 19 (2011) 111–121.
[9] M. Matsuoka, J. Segawa, Y. Makita, S. Ohmachi, T. Kashima, K.I. Nakamura, M.
Hattori, M. Kitano, M. Kise, J. Heterocyclic Chem. 34 (1997) 1773–1779.
[10] J.H. Choi, E.B. Choi, C.S. Pak, Synlett 2 (2003) 166–172.
Acknowledgments
Financial support from the Swedish board for Scientific research
(VR), the Knut and Alice Wallenberg Foundation, the research