1-Sulfonyl-6-Piperazinyl-7-Azaindoles as potent and pseudo-selective 5-HT6 receptor antagonists

Article abstract of DOI:10.1016/j.bmcl.2016.04.024

A series of 1-Sulfonyl-6-Piperazinyl-7-Azaindoles, showing strong antagonistic activity to 5-HT₆ receptor (5-HT₆R) was synthesized and characterized. The series was optimized to reduce activity on D₂ receptor. Based on the selectivity against this off-target and the analysis of the ADME-tox profile, compound 1c was selected for in vivo efficacy assessment, which demonstrated procognitive effects as shown in reversal of scopolamine induced amnesia in an elevated plus maze test in mice. Compound 3, the demethylated version of compound 1c, was profiled against a panel of 106 receptors, channels and transporters, indicating only D₃ receptor as a major off-target. Compound 3 has been selected for this study over compound 1c because of the higher 5-HT₆R/D₂R binding ratio. These results have defined a new direction for the design of our pseudo-selective 5-HT₆R antagonists.

Full text of DOI:10.1016/j.bmcl.2016.04.024

Bioorganic & Medicinal Chemistry Letters 26 (2016) 2610–2615
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
1-Sulfonyl-6-Piperazinyl-7-Azaindoles as potent and pseudo-
selective 5-HT receptor antagonists
6
a
b
b
a
b
Charles-Henry Fabritius , Ullamari Pesonen , Josef Messinger , Raymond Horvath , Harri Salo ,
a
a
a
a
Michał Gał e˛ zowski , Mariusz Galek , Klaudia Stefa n´ ska , Joanna Szeremeta-Spisak ,
Marta Olszak-Płachta , Anna Buda , Justyna Adamczyk , Marcin Król , Peteris Prusis ,
Magdalena Sieprawska-Lupa , Maciej Mikulski , Katja Kuokkanen , Hugh Chapman ,
Radosław Obuchowicz , Timo Korjamo , Niina Jalava , Mateusz Nowak
a
a
a
a
b
a
a
b
b
a
b
b
a,
⇑
a
Selvita S.A., Ul. Bobrzy n´ skiego 14, 30-348 Krakow, Poland
Orion Corporation, Orion Pharma, Orionintie 1A, 02200 Espoo, Finland
b
a r t i c l e i n f o
a b s t r a c t
Article history:
Received 19 February 2016
Revised 7 April 2016
Accepted 10 April 2016
Available online 23 April 2016
A series of 1-Sulfonyl-6-Piperazinyl-7-Azaindoles, showing strong antagonistic activity to 5-HT6 receptor
5-HT R) was synthesized and characterized. The series was optimized to reduce activity on D receptor.
Based on the selectivity against this off-target and the analysis of the ADME-tox profile, compound 1c
was selected for in vivo efficacy assessment, which demonstrated procognitive effects as shown in rever-
sal of scopolamine induced amnesia in an elevated plus maze test in mice. Compound 3, the demethy-
(
6
2
lated version of compound 1c, was profiled against
transporters, indicating only D receptor as a major off-target. Compound 3 has been selected for this
study over compound 1c because of the higher 5-HT R/D R binding ratio. These results have defined a
new direction for the design of our pseudo-selective 5-HT R antagonists.
Ó 2016 Elsevier Ltd. All rights reserved.
a panel of 106 receptors, channels and
Keywords:
Serotonin receptor
3
6
2
5
6
-HT receptor
Antagonists
Azaindole
6
Cognitive impairment
Novel therapies against dementias and in particular Alzheimer’s
disease (AD) constitute one of the biggest medical needs in devel-
oped countries. With the morbidity reaching over 35 million cases
One of the approaches to cognitive improvement is blocking the
serotonin 5-HT receptor with an antagonist. 5-HT R is expressed
almost exclusively in the central nervous system in humans,
mainly in hippocampus, striatum and nucleus accumbens. 5-HTRs
couple to Gs-protein and stimulate adenylyl cyclase activity.
6
6
1
worldwide and the annual worldwide cost of above US$315 bil-
2
lion this group of diseases remains in the center of interest of
pharmaceutical industry and drug research around the world. Dis-
ease modifying treatments do not yet exist, and disadvantages of
current standard symptomatic medications, namely inhibitors of
acetyl cholinesterase (e.g., donepezil, rivastigmine, galantamine)
and NMDA receptors (e.g., memantine) have poor tolerability,
low efficacy and present challenges with patient compliance
because of suboptimal dosing regimens and side effects (mainly
gastrointestinal). These shortcomings result in symptomatic thera-
Antagonism of 5-HT
mance in rodents in numerous memory related tasks.
tantly, these effects seem to be translated to humans—Lundbeck
has recently announced positive results from a Phase II clinical trial
of its selective 5-HT
6
R was shown to improve cognitive perfor-
5–9
Impor-
6
R antagonist (idalopirdine; Lu AE58054) in
improving cognitive performance in mild to moderate AD
1
0
11
patients
5-HT R provides a potential therapeutic strategy for cognitive
symptoms of schizophrenia and obesity. The mode of action
of 5-HT R antagonists has been elucidated in vivo by means of
electrophysiology and microdialysis where it was shown that
antagonizing 5-HT R enhances glutamatergic, cholinergic and
monoaminergic neurotransmission.
Comprehensive reviews of 5-HT
were published in recent years by Holenz, Liu, Ivachtchenko
and initiated Phase III.
Additionally, antagonizing
6
3
12
13
pies which benefit the patient for only about one year on average,
whereas the potential need for therapy for AD patients can last for
even 8.5 years (considering the duration from the onset of the dis-
6
1
4
15
4
ease to the severe stage). Due to the above mentioned reasons an
6
innovative approach to enhance cognition in AD patients is
desirable.
6
R related medicinal chemistry
1
6
17
18
1
9
and Lopez-Rodriguez. According to Lopez-Rodriguez most of
the known 5-HT R ligands can be clustered into four structural
families taking into account the groups that occupy the main
⇑
Corresponding author. Tel.: +48 12 297 47 00; fax: +48 12 297 47 01.
E-mail address: mateusz.nowak@selvita.com (M. Nowak).
6
http://dx.doi.org/10.1016/j.bmcl.2016.04.024
960-894X/Ó 2016 Elsevier Ltd. All rights reserved.
0

C.-H. Fabritius et al. / Bioorg. Med. Chem. Lett. 26 (2016) 2610–2615
2611
pharmacophoric features: bisarylsulfonamides, indoles, indole-like
derivatives and non-sulfonyl compounds. Those key structural ele-
ments for 5-HT6 antagonism can be modeled into a simplified
in the presence of N,N-Diisopropylethylamine to afford 4-chloro-6-
(N-substituted-piperazynyl)-1H-7-azaindoles 5 using the proce-
2
2
dure of Reissert-Henze reaction. In order to study the influence
1
9,20
1
pharmacophore:
a positive ionizable atom, an aromatic ring-
of the aryl group R , compound 5 was reacted with isobutyl sul-
hydrophobic site, a hydrogen bond acceptor and a hydrophobic
site. Some diversified atypical compounds have also been reported,
although they share common structural part of those 4 main fam-
ilies. According to the recent review of Ivachtchenko, which is
focused mainly on the selectivity profile of 5-HT ligands, these
6
can be further classified into three categories: multimodal/multi
target, pseudo-selective and selective.
fonyl chloride in the presence of sodium hydride, followed by a
Suzuki coupling with the corresponding boronic acid to give com-
pounds 1a–p. Similarly to understand the role of R on the sulfone,
2
compounds 2a–f were synthesized by exchanging the two previous
steps. First, the Suzuki coupling was performed with 4-trifluo-
romethylphenyl boronic acid and then sulfonylation reaction was
performed with the appropriate sulfonyl chloride. Demethylated
compound 3 was obtained by following the same synthetic path-
way using N-Boc-piperazine. An additional deprotection step with
trifluoroacetic acid was needed to obtain the final compound.
All compounds were screened in search of high affinity
Here we report the discovery and pharmacological characteri-
zation of a series of 6-piperazinyl-7-azaindoles bearing an alkyl
or hetero alkyl sulfone in position 1 and an aryl or hetero aryl in
position 4 (compounds 1a–p, 2a–f and 3 represented in Scheme 1).
Those compounds could be assigned to the indole-like family:
azaindole is used here as a bioisostere of the typical indole core.
We also synthesized the bisarylsulfonamides derivatives on this
(K
dopamine D
The affinities (K
i
620 nM) on the 5-HT
receptor (K P200 nM) in a radioligand binding assay.
2
) of the studied compounds for the 5-HT R and D R
6
R and low to negligible affinity to the
2
i
i
6
2
1
3
azaindole core but those compounds were not selected for fur-
ther studies due to the superior results obtained for the aliphatic
analogs (results not reported) and tolerance of this new chemical
series for non-lipophilic moieties. Furthermore, the removal of
the sulfonyl group is not tolerated and results in losing the potency
were determined indirectly by displacement of [ H]-LSD and
3
[ H]-NMSP, respectively. Results are displayed in Tables 1 and 2.
1
2
Our first compound 1a (R = Ph and R = isobutyl) was found to
have potent in vitro binding affinity toward 5-HT R (K = 16 nM)
R binding ratio of 35. Initially, we
6
i
and an acceptable 5-HT
6
R/D
2
1
for 5-HT
6
receptor.
were interested to know if R = Ph was the optimal group in this
chemical series. So we investigated the influence of the group R1
(compounds 1a–p, Table 1). A scan through various substitutions
on the phenyl was initially tested (compounds 1a–j, Table 1). Com-
pound 1b with an ethyl substituent in position 4 was shown to
Most compounds from these series are potent and pseudo-
selective according to the classification of Ivachtchenko. In our pro-
ject we have defined D receptor as an antitarget. This receptor is
2
connected to extrapyramidal symptoms most commonly caused
by typical antipsychotics. Additionally, the selectivity against a
broad panel of diverse targets was checked for the selected
representative example 3. Full SAR for the series is available for
have similar potency and 5-HT
ethyl was replaced with –CF , compound 1c was 3 times more
potent. Although the compound had higher affinity for D R, the
5-HT R/D R binding ratio was the best one among the three afore-
mentioned compounds.
Next, we have tested effect of polar substituents on 5-HT
R affinities of the compounds. Interestingly, compound 1f with –
OMe was the most potent 5-HT R binder (K = 2nM). However, the
5-HT R/D R binding ratio was 10 fold lower than our prototype
compound 1c. Replacement of –OMe by –OBn (compound 1g)
6 2
R/D R binding ratio to 1a. When
3
2
5
-HT
As shown in Scheme 1, the lead structure template was divided
into two structural regions for analog optimization, the aryl group
6
and D
2
receptors.
6
2
6
R and
1
2
R
and the pendant alkyl R . The synthesis of the proposed com-
D
2
pounds was achieved as described in Scheme 1. 4-Chloro-7-azain-
dole 4 was transformed into N-oxide with m-CPBA.22 N-Oxide 4
was alkylated with dimethyl sulfate and the obtained intermediate
was treated with commercially available N-substituted piperazines
6
i
6
2
reversed the binding ratio toward D
2
R by a factor of 30. Compound
2
3
1
R1
4
N
Cl
NH
Cl
N
S
O
S
O
O
O
Cl
NH
a
N
b
N
c
N
N+
N
N
N
N
N
N
O-
4
3
R
a: R3 = Me, 32%
1a-p: 5%-81%
5
5
6: 75%
3
b: R = Boc, 34%
c
2
2
3
F
3
C
F
3
C
3
3
F C
R2
O
1
1
NH
4
N
S
4
N
S
O
O
O
d
b
N
N
N
N
N
N
N
R
N
R
N
H
3
3
7
7
a: R3 = Me, 52%
2a-f: R3 = Me, 30%-59%
3: 67%
3
8: R = Boc, R2 = iBu, 49%
3
b: R = Boc, 71%
Scheme 1. Reagents and conditions: (a) Me
2
SO
4
, MeCN, piperazine derivatives; 60 °C, 6 days; (b) R
2 2
SO Cl, NaH, dry DMF; 0 °C, 90 min; (c) boronic acid or boronic ester, Pd
(
OAc) , S-Phos, K PO , toluene; 130 °C, 24 h; (d) CF
2
3
4
3
CO H, DCM; RT, 1 h. For the definition of R1 and R2, see Tables 1 and 2.
2

2
612
C.-H. Fabritius et al. / Bioorg. Med. Chem. Lett. 26 (2016) 2610–2615
Table 1
1i with –CN showed comparable potency as compound 1c. How-
6 2
SAR and 5-HT R/D
R binding ratio profile: analysis of R1
ever, this compound was 3 times less selective. The same was
observed for a bicyclic version (compound 1k bearing an indole
substituted in position 5). Finally, the conversion of this –CN group
to –Amide led to completely inactive compound (compound 1j).
We then investigated the direct replacement of the phenyl
group by a heteroaryl such as the pyridine 1l or pyridone 1m.
Either of the compounds did not show any improvement. We also
tried 5-membered ring analogs like pyrazole 1n, thiophene 1o,
isoxazole 1p. The most promising derivative in this set was the
R1
N
S
O
O
N
N
N
R1
5-HT
nM)
a
D
(nM)
R K
a
Compound
6
R K
i
2
i
6 2
Ratio 5-HT R/D R
thiophene analog with respect to affinity and 5-HT
ratio.
6 2
R/D R binding
(
1
1
1
1
1
1
1
1
1
1
a
b
c
d
e
f
g
h
i
Ph
16
11
4
48
24
2
290
22
7
547
614
367
630
113
22
86
108
205
35
56
80
13
5
9
0.3
5
We also investigated variations on R2 to determine if isobutyl
was the optimum alkyl or heteroalkyl on the sulfone. This SAR
C
C
C
C
C
C
C
C
C
6
H
6
H
6
H
6
H
6
H
6
H
6
H
6
H
6
H
4
4
4
4
4
4
4
4
4
(4-Et)
(4-CF
(3-CF
(2-CF
3
)
)
)
2
analysis is described in Table 2. When R = Me was tested, loss of
3
3
potency and selectivity was observed compared to compound 1c.
(4-OMe)
(4-OBn)
(4-OH)
(4-CN)
The removal of one –CH
c gave us compound 2b. This was sufficient to get similar activity,
unfortunately in detriment of the selectivity over D R (K = 25 nM).
2
unit to the isobutyl group of compound
1
29
-
2
i
2
j
(4-CONMe
2
)
>2000
>2000
The activity was 4 fold lower when a more rigid group R = cyclo-
hexyl was tried. Addition of polar substituents into the alkyl
groups lead to compounds 2d and 2e, which did not provide
improvements towards the desired overall profile.
H
N
1
k
l
5
382
76
N
The demethylated version of compound 1c was also synthe-
1
36
>2000
>55
sized. The activity towards 5-HT
as it is shown in Figure 1. Interestingly, compound 3 gave better
5-HT R/D R binding ratio (567x). In order to confirm the antago-
nism on the 5-HT receptor, the functional activity through mea-
6
R was similar for both compounds
HO
N
1
1
m
n
231
87
>2000
>2000
>8
6
2
6
N
surement of cAMP levels in a cell-based assay was determined in
Cerep. Kb value for compound 1c is given in Figure 1.
N
>23
The selectivity of compound 3 was tested at a concentration of
1 lM against a Cerep panel consisting of 106 neurotransmitter
receptors, transporters and ion channels (Fig. S1). Compound 3
has been selected for this study over compound 1c because of
6 2
the higher 5-HT R/D R binding ratio.
1
1
o
p
S
12
640
53
N
O
129
>2000
>15
According to the new classification proposed recently by
Ivachtchenko, compound 3 might be classified as pseudo-selective
a
À6
À13
Compounds at eight concentrations ranging from 5 Â 10 to 5 Â 10
M were
(
6
the compound shows a ratio of its affinity to the 5-HT R compared
tested in triplicates. Data analysis was performed in GraphPad Prism software and
presented in the form of graphs on semi-log scale. Curve fitting was done using
nonlinear regression using one site competition model.
1
8
to other biological targets affinities of less than 250). It is worth
noting though, that most of the off–targets detected for compound
3
(
show radioligand displacement only slightly above the threshold
50%) also used in Ivachtchenko classification to define weak inter-
actions. 60% radioligand displacement at 1 M for the D receptor
in the Cerep panel corresponds to K = 1.7 M D R measured in our
in-house assay. Only a single off-target (D receptor) can be treated
as equipotent to 5-HT R (94% vs 100% for 5-HT6). Strong radioli-
l
2
i
l
3
2
Table 2
6
6 2
SAR and 5-HT R/D
R binding ratio profile: analysis of R2
gand displacement (72%) has also been observed for 5-HT1B recep-
tor. Full results of Cerep panel have been included in Supporting
information (Fig. S1).
To complement medicinal chemistry approach and gain some
insight of the molecular interactions that could influence affinity
3
F C
R2
O
N
S
O
N
6
and selectivity towards 5-HT R, we have performed in silico
N
N
F
3
C
3
F C
a
a
Compound
R2
5-HT
nM)
6
R K
i
D
(nM)
2
R K
i
Ratio 5-HT
6
R/D
2
R
N
S
O
N
S
(
O
O
O
N
N
1c
2a
2b
2c
Isobutyl
Me
iPr
4
54
6
367
1231
25
80
22
4
>133
N
N
N
N
H
Cyclohexyl
15
>2000
2
d
e
7
8
>2000
>285
6
1c
6
3
6
O
O
Ki 5-HT
Ki D
Kb 5-HT
R = 4 nM
Ki 5-HT R = 3 nM
2
R = 367 nM
R = 250 nM
Ki D R = 1702 nM
2
2
52
6
a
See footnote a of Table 1.
Figure 1. Comparison of compounds 1c and 3 (demethylated piperazine).

C.-H. Fabritius et al. / Bioorg. Med. Chem. Lett. 26 (2016) 2610–2615
2613
Figure 2. Compound 1c docked to the 5-HT
6
R. Key interaction residues of the receptor are shown in the stick representation. The view of the receptor from the extracellular
side; extracellular loop 2 (ECL2) and a part of the transmembrane helix 5 (TM5) are removed for clarity.
flexible docking of studied compounds in homology models of
-HT and D receptors.
All studied compounds were docked to the selected receptor
models and binding poses were carefully analyzed. The predomi-
nantly observed binding mode is similar in both 5-HT R and D R.
The compounds bind with the piperazine—core plane perpendicu-
selectivity ratio in the range of 4 to 22, which rises to several hun-
2
5
6
2
dred (compounds are effectively inactive for D
2
R) when R is bulky.
Limited volume of the upper vestibule may also influence selec-
1
1
tivity of the aromatic R substituent. Aromatic binding pocket of R
6
2
moiety is closed at the bottom by a rigid lid formed by Phe 6.44
residue. This residue is conserved in GPCRs and forms ligand inter-
1
2
1
lar to the long axis of a receptor and R –R plane parallel to the
long axis of the receptor (Fig. 2).
actions in rhodopsin. As the bottom of the R pocket is rigid com-
1
pounds with longer R fragments are shifted upwards toward the
1
In this binding mode R substituent binds deep between helices
vestibule. However, they can easier accommodate in the larger ves-
2
TM3, TM5 and TM6, while R binds in the vestibule formed by the
6 2
tibule of 5-HT R rather than D R. This explains the trend observed
1
cytoplasmic ends of TM helices. Compounds are stabilized in the
by SAR analysis, where compounds with longer R are more selec-
binding site of 5-HT
gen atom of the piperazine moiety and Asp 3.32,
the azaindole core with Phe 6.52, stacking of the R aromatic
6
R by a salt bridge between the charged nitro-
tive. For example, among the compounds 1c–1e, selectivity
p–p
stacking of
decreases going from the longest p-CF
3
-Ph to the shortest
1
p
–p
o-CF -Ph. Other compounds also follow this trend.
3
moiety with Trp 6.48 and hydrogen bonds between sulfone group
Compounds 1c and 3 were further profiled in ADME-toxicity
assays to investigate potential liabilities linked to this particular
type of scaffold. Both compounds had excellent permeability
through the PAMPA membrane. The IC50 values for CYP 3A4 and
CYP 2D6 were found to be better for compound 1c than 3, indicat-
ing less likely drug–drug interactions in humans through the major
metabolic enzymes tested. Compound 1c had acceptable in vitro
metabolic stability in the three species human/mouse/rat as shown
by the calculated intrinsic clearance. hERG liability was also evalu-
ated using an automated patch clamp assay. The activities towards
hERG were similar for both compounds as it is shown in Table 3.
Kinetic solubility was also determined for the free base of com-
and Ser 5.43 and predominantly Asn 6.55 (Fig. 2). Ligand–receptor
2
3,24
interactions are in agreement with the available literature data
2
5
and with the results of other docking studies for 5-HT
In the next step, docking studies were used to explain 5-HT
versus D R selectivity for the studied compounds. Generally,
6
R.
6
R
2
target-based investigations of selectivity are difficult for similar
targets as affinity differences depend not only on the different
shape of the binding site but also on the different hydration pattern
of the binding site, different stabilities and free energies of water
molecules in the active site, changes in general flexibility of the
binding site and many more effects. Here, we analyzed differences
in the sequence and shape of the binding sites of both studied
receptors which may impact proposed binding mode. However, it
should be noted that other factors, mentioned above, may con-
tribute to a large extent to the observed selectivity of the studied
compounds. Comparison of sequence and shape pointed to posi-
pound 1c and showed to be a poorly soluble compound (8 lg/mL).
A study of different salt form has been performed. As shown in
Table 4, solubility was slightly improved when compound 1c was
converted to the hydrobromide salt.
Compound 1c was selected for in vivo cognitive testing in
mouse based on its biology and ADME-tox profile.
We used the Elevated Plus Maze (EPM) with scopolamine as an
6
in vivo screening test to assess the effects of 5-HT R antagonism in
mice. Briefly, mice were administered with the test compound, and
10 min later with scopolamine. In the acquisition phase (T1) the
time spent in the open arm (OAT%; [open arm time/open
+ enclosed time] * 100) and the latency of entering the open arm
tion 6.55, which is occupied by Asn in 5HT
chains of the residues in this position project to the upper vestibule
and His 6.55 in D R is shifted upwards compared to Asn 6.55 in
-HT R. As a consequence His 6.55 occupies considerably more
6 2
R and His in D R. Side
2
5
6
2
space, effectively closing the upper vestibule for bulky R sub-
stituents (Fig. 3). This structural observation is confirmed by SAR,
where compounds 2a and 2b with small R2 substituent have low

2
614
C.-H. Fabritius et al. / Bioorg. Med. Chem. Lett. 26 (2016) 2610–2615
Figure 3. Superimposition of 5-HT
6
R (green) and D
2
R (pink) receptors. For residues in position 6.55 (stick representation) molecular surface is shown. Compound 1c and key
residues of the binding pocket responsible for ligand binding are shown as in Figure 2. The view of the receptor from the side. Extracellular loop 2 (ECL2) and part of the
transmembrane helix 5 (TM5) removed for clarity.
Table 3
(TL) were recorded. Mice prefer the less aversive closed compart-
ments, and learn quickly to go there, if the spatial learning and
memory are not disturbed with amnestic compound, like mus-
carinic receptor antagonist scopolamine. In the retention phase
(T2) 24 h after the T1 phase decrease in OAT% and TL is believed
to indicate learning and memory. The scopolamine treatment
increased the OAT% and TL compared to vehicle treatment in both
T1 and T2 phases which can be considered as disturbance of mem-
ory. The memory impairment could be prevented with the treat-
ment of optimal low dose of compound 1c as measured with
OAT% (Fig. 4). Even if there was no clear dose response either in
the acquisition or the retention phase, it can be concluded that
compound 1c showed in vivo efficacy with a narrow dose range
in our in vivo screening model. These U-shaped dose response
curves are common for nootropic compounds.
ADME-tox properties for selected compounds
Assays
Compound 1c
Compound 3
PAMPA (nm/s)
537
920
CYP (
Clint (
hERG, IC50 (
l
M)
L/min/mg protein)
M)
3A4: 1.1 <IC50 <3.3
3A4: IC50 <1.1
2D6: IC50 >10
—
2
D6: 3.3 <IC50 <10
h/m/r
0/63/55
l
2
l
2.1
1.4
Table 4
Solubility study of compound 1c
Salt
Free
base
HCl HBr
H
2
SO
4
Fumaric/tartaric/
citric acid
Kinetic solubility
8
14
65
10
20/11/14
Thus we report a new series of potent and pseudo-selective
(l
g/mL)
6
5HT receptor antagonists with procognitive efficacy in a mouse
model of spatial learning and memory. Many of the tested
Figure 4. Total percentage of the time spent in the open arms (OAT%) and the transfer latency (TL) to the enclosed arm in the elevated plus maze after subcutaneous
administration of compound 1c at different doses. Data are presented as mean ± SEM and median (n = 13–15). White bar is the control group, black bar is the scopolamine
(
SCO) treated control group, and shaded bars are scopolamine treated test groups with different concentrations of compound 1c. + significantly different from the vehicle
group, # significantly different from the scopolamine group, ⁄ T2 value significantly different from the respective T1 value of the group. All values measured after 24 h (phase
T2) of acutely measured phase T1 values were lower reflecting spatial learning and memory. One symbol p <0.05, three symbols p <0.001.

C.-H. Fabritius et al. / Bioorg. Med. Chem. Lett. 26 (2016) 2610–2615
2615
compounds showed high activities in the nanomolar range with
HT R/D R binding ratio over 100 for some compounds. Computa-
4. Francis, P. T.; Palmer, A. M.; Snape, M.; Wilcock, G. K. J. Neurol. Neurosurg.
Psychiatry 1999, 66, 137.
5
6
2
5.
King, M. V.; Sleight, A. J.; Woolley, M. L.; Topham, I. A.; Marsden, C. A.; Fone, K.
tional analysis helped us to rationalize observed SAR in terms of
potency and selectivity and facilitated further development of
compounds and optimization of ligand–receptor interactions in
C. Neuropharmacology 2004, 47, 195.
6. Foley, A. G.; Murphy, K. J.; Hirst, W. D.; Gallagher, H. C.; Hagan, J. J.; Upton, N.;
Walsh, F. S.; Regan, C. M. Neuropsychopharmacology 2004, 29, 93.
7
8
.
.
Loiseau, F.; Dekeyne, A.; Millan, M. J. Psychopharmacology (Berl) 2008, 196, 93.
Perez-García, G.; Meneses, A. Pharmacol. Biochem. Behav. 2005, 81, 673.
1
2
both R and R regions.
Optimization of compounds was driven not only by the
improvement of on-target activity and 5HT R/D R binding ratio
6 2
but also by in vitro ADME and safety parameters. Due to the most
favorable balance between potency, selectivity, safety properties
hERG and CYP inhibition) and ADME properties (solubility, PAMPA
and metabolic stability) compound 1c was selected for further
in vivo PK and efficacy studies. In the mouse model of spatial learn-
ing (EPM) experiment, compound 1c prevented the amnestic effect
of scopolamine in the memory acquisition (learning) and 24 h later
in the retention phase (memory).
9. Hatcher, P. D.; Brown, V. J.; Tait, D. S.; Bate, S.; Overend, P.; Hagan, J. J.; Jones, D.
N. Psychopharmacology (Berl) 2005, 181, 253.
1
1
0. Wilkinson, D.; Windfeld, K.; Colding-Jørgensen, E. Lancet 2014, 13, 1092.
1. http://investor.lundbeck.com/releasedetail.cfm?ReleaseID=750931, accessed
on 2015-08-24.
1
2. Arnt, J.; Bang-Andersen, B.; Grayson, B.; Bymaster, F. P.; Cohen, M. P.; DeLapp,
N. W.; Giethlen, B.; Kreilgaard, M.; McKinzie, D. L.; Neill, J. C.; Nelson, D. L.;
Nielsen, S. M.; Poulsen, M. N.; Schaus, J. M.; Witten, L. M. Int. J.
Neuropsychopharmacol. 2010, 13, 1021.
(
1
1
3. Heal, D.; Gosden, J.; Smith, S. Int. Rev. Neurobiol. 2011, 96, 73.
4. Tassone, A.; Madeo, G.; Schirinzi, T.; Vita, D.; Puglisi, F.; Ponterio, G.; Borsini, F.;
Pisani, A.; Bonsi, P. Neuropharmacology 2011, 61, 632.
15. Hirano, K.; Piers, T. M.; Searle, K. L.; Miller, N. D.; Rutter, A. R.; Chapman, P. F.
Life Sci. 2009, 84, 558.
1
6. Holenz, J.; Pauwels, P. J.; Díaz, J. L.; Mercè, R.; Codony, X.; Buschmann, H. Drug
Supplementary data
Discovery Today 2006, 11, 283.
17. Liu, K. G.; Robichaud, A. J. Drug Dev. Res. 2009, 70, 145.
18. Ivachtchenko, A. V.; Ivanenkov, Y. A. Curr. Bioact. Compd. 2013, 9, 64.
19. Benhamú, B.; Martín-Fontecha, M.; Vázquez-Villa, H.; Pardo, L.; López-
Rodríguez, M. L. J. Med. Chem. 2014, 57, 7160.
Supplementary data (experimental protocols for all compounds,
-HT R binding, D R binding, Cerep 5-HT functional, Cerep diver-
6 2 6
5
sity panel, automated patch clamp IhERG, PAMPA, CYP 3A4/2D6
inhibition, metabolic stability, kinetic solubility assays, PK study,
elevated plus maze test and homology modeling of 5HT and D
6 2
20. Holenz, J.; Mercè, R.; Díaz, J. L.; Guitart, X.; Codony, X.; Dordal, A.; Romero, G.;
Torrens, A.; Mas, J.; Andaluz, B.; Hernández, S.; Monroy, X.; Sánchez, E.;
Hernández, E.; Pérez, R.; Cubí, R.; Sanfeliu, O.; Buschmann, H. J. Med. Chem.
2005, 48, 1781.
receptors) associated with this article can be found, in the online
version, at http://dx.doi.org/10.1016/j.bmcl.2016.04.024.
21. Dunn, R.; Nguyen, T. M.; Xie, W.; Tehim, A.; W.O. Patent 101247, 2008.
22. Storz, T.; Bartberger, M. D.; Sukits, S.; Wilde, C.; Soukup, T. Synthesis 2008, 2,
201.
23. Hirst, W. D.; Abrahamsen, B.; Blaney, F. E.; Calver, A. R.; Aloj, L.; Price, G. W.;
Medhurst, A. D. Mol. Pharmacol. 2003, 64, 1295.
References and notes
2
4. Harris, R. N., 3rd; Stabler, R. S.; Repke, D. B.; Kress, J. M.; Walker, K. A.; Martin,
R. S.; Brothers, J. M.; Ilnicka, M.; Lee, S. W.; Mirzadegan, T. Bioorg. Med. Chem.
Lett. 2010, 20, 3436.
1
2
.
.
Querfurth, H. W.; LaFerla, F. M. N. Engl. J. Med. 2010, 362, 329.
http://www.alz.co.uk/research/files/WorldAlzheimerReport-
ExecutiveSummary.pdf, accessed on 2015-08-24.
25. Pullagurla, M. R.; Westkaemper, R. B.; Glennon, R. A. Bioorg. Med. Chem. Lett.
2004, 14, 4569.
3
.
Sun, Y.; Lai, M. S.; Lu, C. J.; Chen, R. C. Eur. J. Neurol. 2008, 15, 278.