Potent dihydroquinolinone dopamine D2 partial agonist/serotonin reuptake inhibitors for the treatment of schizophrenia

Article abstract of DOI:10.1016/j.bmcl.2010.02.105

A dihydroquinolinone moiety was found to be a potent serotonin reuptake inhibitor pharmacophore when combined with certain amines. This fragment was coupled with selected D₂ ligands to prepare a series of dual acting compounds with attractive in vitro profiles as dopamine D₂ partial agonists and serotonin reuptake inhibitors. Structure-activity studies revealed that the linker plays a key role in contributing to D₂ affinity, function, and SRI activity.

Full text of DOI:10.1016/j.bmcl.2010.02.105

Bioorganic & Medicinal Chemistry Letters 20 (2010) 2983–2986
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
Potent dihydroquinolinone dopamine D2 partial agonist/serotonin
reuptake inhibitors for the treatment of schizophrenia
a
a,
b
b
b
Yinfa Yan a, , Ping Zhou , David P. Rotella , Rolf Feenstra , Chris G. Kruse , Jan-Hendrik Reinders ,
Martina van der Neut b, Margaret Lai c, Jean Zhang c, Dianne M. Kowal c, Tikva Carrick c, Karen L. Marquis c,
Mark H. Pausch c, Albert J. Robichaud a
*
*
a Chemical Sciences, Wyeth Research, CN 8000 Princeton, NJ 08543, United States
b Solvay Pharmaceuticals Research Laboratories, C.J. van Houtenlaan 36, 1381 CP, Weesp, The Netherlands
c Discovery Neuroscience, Wyeth Research, CN 8000 Princeton, NJ 08543, United States
a r t i c l e i n f o
a b s t r a c t
Article history:
A dihydroquinolinone moiety was found to be a potent serotonin reuptake inhibitor pharmacophore
when combined with certain amines. This fragment was coupled with selected D2 ligands to prepare a
series of dual acting compounds with attractive in vitro profiles as dopamine D2 partial agonists and sero-
tonin reuptake inhibitors. Structure–activity studies revealed that the linker plays a key role in contrib-
uting to D2 affinity, function, and SRI activity.
Received 28 January 2010
Revised 25 February 2010
Accepted 26 February 2010
Available online 3 March 2010
Ó 2010 Elsevier Ltd. All rights reserved.
Keywords:
Serotonin reuptake inhibitor
Dopamine D2 partial agonist
Schizophrenia
Based on the hypothesis that the efficacy of antipsychotic
agents can be expanded by combining a dopamine D2 partial ago-
nist with serotonin reuptake inhibition (SRI),1–3 we reported a ser-
ies of tetrahydrocarbazole-based dual acting compounds that
showed antipsychotic-like activity in animal models (see struc-
tures 1 and 2 in Fig. 1).4 The receptor binding profile of these mol-
ecules indicated less than a 10-fold separation between affinity at
the primary targets (SRI and D2) and potentially undesirable recep-
The design hypothesis for compounds described in this report is
embodied in a compound template (see Fig. 2) which includes a D2
ligand, an SRI moiety, and a linker. Selection of the dihydroquinoli-
none moiety as the SRI pharmacophore was based in part on the
observation that aripiprazole, 3 (Fig. 2) an approved agent for the
treatment of schizophrenia displays weak SRI activity (Ki 1080 nM
using a citalopram binding assay similar to that used in this work).7,8
The choice of D2 pharmacophores was guided by dopamine D2 moi-
eties known in the literature.9,10 The two building blocks can be effi-
ciently linked via alkyl- and piperidine-based linkers.
tors such as a
1.5,6 In an attempt to address this deficiency and bet-
ter understand the complex structure–activity relationships that
determine primary target selectivity, we now report the results ob-
tained with additional SRI and D2 building blocks.
The synthesis of the final targets is outlined in Schemes 1–5. The
oxygen linked compounds were prepared as outlined in Scheme 1.11
H
O
N
N
O
F
NH
N
F
N
H
O
NH
N
H
N
H
O
1
2
Figure 1. D2/SRI partial agonists.
* Corresponding authors. Tel.: +1 215 801 0191.
E-mail address: drotella@comcast.net (D.P. Rotella).
0960-894X/$ - see front matter Ó 2010 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2010.02.105

2984
Y. Yan et al. / Bioorg. Med. Chem. Lett. 20 (2010) 2983–2986
H
O
N
O
H
Cl
N
L
N
O
Cl
N
SRI Pharmacophore
L: Linker
D2
D2 Pharmacophore
3 aripiprazole
4
Figure 2. D2/SRI design approach.
NH2
BocN
a
HO
N
H
O
c
O
N
H
O
b
OH
5
e
7
6
HN
d
9 (n = 1)
Br
10 (n = 2)
O
N
O
O
N
H
O
n
H
8
Scheme 1. Synthesis of dihydroquinolinone analogs I. Reagents and conditions: (a) ClCH2CH2COCl, NaHCO3/MeOH/H2O/rt, 24 h; (b) AlCl3 neat/200 °C; (c) t-butyl 4-(2-
nitrophenylsulfonyloxy)piperidine-1-carboxylate, Et3N/CH2Cl2; (d) HCl/ether; (e) Br(CH2)nBr, Cs2CO3/acetone/DMF reflux.
Br
Br
Boc
N
O
O
P
a
b
OEt
OEt
O2N
CH3
O2N
c
11
12
O
OCH3
e
H3CO
NBoc
O
d
O2N
13
NH
NBoc
f
O
N
O
N
H
H
15
14
Scheme 2. Synthesis of dihydroquinolinone analogs II. Reagents and conditions : (a) NBS and AIBN/CCl4 reflux 24 h; (b) (POEt)3 160 °C, 3 h; (c) NaH/DMSO rt; (d) Pd(OAc)2,
di-isopropyl ethyl amine/DMF; (e) Pd/C, H2/MeOH 6 h; (f) HCl/ether.
Acylation of 3-aminophenol with 3-chloropropanoyl chloride affor-
ded the desired product which was cyclized under Friedel–Crafts
conditions to give lactam 6, followed by alkylation and acidic depro-
tection to give compound 8. Alkylation of 6 with 1,2-dibromoethane
or 1,3-dibromopropane affords intermediates 9 and 10, respec-
tively. Scheme 2 shows the synthesis of piperidinylmethyl 3,4-dihy-
droquinolinone 15.11 Bromination of 3-nitro-4-bromotoluene 11
with NBS affords 2-bromo-4-(bromomethyl)-1-nitrobenzene,
which was converted to phosphonate 12. Wittig–Horner reaction
of 12 yielded the corresponding olefin intermediate, followed by
Heck coupling with methyl acrylate, catalytic hydrogenation, and
deprotection furnished 15. The piperidinylmethyl 3,4-dihydroquin-
olinone 15 was coupled with various D2 moieties to yield the desired
products by either direct alkylation (halide) or reductive amination
(ketones or aldehydes). Several intermediates were prepared by
similar chemistry utilizing different ketones (Scheme 3).
Br
O
OEt
P
O
N
O2N
OEt
Boc
N
O
12
NH
NH
O
N
O
N
H
H
17
16
Scheme 3. Synthesis of dihydroquinolinone analogs III.

Y. Yan et al. / Bioorg. Med. Chem. Lett. 20 (2010) 2983–2986
2985
OCH3
N
a
O
Br
O
P
O
O
OEt
OEt
H3CO
N
O2N
b
O2N
12
18
NH
c
O
N
H
H
16
~3:1 mixture of isomers
Scheme 4. Synthesis of bridged compounds. Reagents and conditions: (a) NaH/DMSO, 0 °C–rt; (b) Pd(OAc)2, di-isopropyl ethyl amine/DMF; (c) Pd/C, H 50 psi/MeOH 6 h.
H
O
O
X
N
O
H
N
X
O
19
N
Di-isopropyl ethyl amine
NaI/1-BuOH 140OC
HN
O
HN
28 (X=O)
25 (X=C)
8 (X=O)
15 (X=C)
Scheme 5. Synthesis of target compounds.
In the case of [3.2.1]azabicyclo fragment 16 (Scheme 3), NMR
analysis showed that this intermediate was obtained as a 3:1 mix-
ture of exo- and endo isomers. These two isomers were separated
by preparative HPLC, and NOE experiments were essential to
determine relative stereochemistry on the azacycle. In the major
exo isomer derived from 16, the axial proton on the cyclohexane
ring displayed NOE interactions with protons in the ethyl bridge.
These interactions were not observed in the endo isomer where
the methylene hydrogens in the axial position showed NOE effects
with protons in the bridge. Similar interactions were observed in
the separate isomers derived from 17. The D2 building blocks used
(19–21) are listed in Figure 3. The synthesis of these compounds
was carried out as previously described in the literature.12,13
Coupling the amino-SRI and D2 fragments was carried out
optimally in refluxing 1-butanol (Scheme 5) in the presence of di-
isopropyl ethyl amine and an iodide source to afford products in
40–78% yield. It was observed that improved yields were obtained
by conversion of the amine hydrochloride to the corresponding free
base by stirring a methanol solution in the presence of a basic ion ex-
change resin (Amberlyst A-26(OH) ion exchange resin).
Conformational restriction of the alkyl amino linker provides
piperidinyl derivatives 25–30 as shown in Table 2. This series
examined the effect of substituting carbon for oxygen at the point
of attachment to the SRI pharmacophore. In general, these piperi-
dine derivatives showed improved SRI affinity compared to alk-
oxy-linked analogs 22–24. When oxygen and carbon linked
compounds are compared, using the same D2 ligands, some carbon
analogs provide more potent SRI affinity (cf. 25 vs 28–30) with
similar D2 affinity. However, this is not a general property, as oxy-
gen analogs 29–30 furnish derivatives with comparable SRI affinity
to methylene-connected compounds 27 and 28. This observation is
similar to that observed earlier in that there appears to be a com-
plex and not completely predictable structure–activity relationship
in this series of dual acting compounds.4
When the piperidine was rigidified with a bicyclic system (8-
(B1) or 3-(B2) azabicyclo[3.2.1]octane), the SRI binding affinity
was improved up to 10-fold (31–33) over piperidine linked com-
pounds (26–30). In this set of analogs, it is noteworthy that linkers
influence SRI binding more than D2 binding and the effect on D2
binding is somewhat variable. The data reported in Table 3 was ob-
tained using the mixtures of exo and endo isomers described above.
Interestingly, the separated exo and endo isomers derived from 31–
34 showed comparable SRI and D2 affinity. Two of these bicyclic
linker compounds (32 and 33) displayed the desired dopamine
D2 partial agonist activity between 0.2 and 0.4. 8-Aza analog 31
Target compounds were screened for binding at the human
serotonin reuptake site in a [3H]citalopram displacement assay.4
D2 receptor binding was measured using [3H]spiperone as the li-
gand.4 Compounds with Ki values less than 50 nM at both receptors
were evaluated in an in vitro D2 functional assay that measured
changes in intracellular cAMP levels.4 Agonist activity was defined
showed full antagonist activity in this screen (
a <0.1), while 3-
as a ratio (a-value) relative to the full agonist quinpirole (
a-va-
aza derivative 34 functioned as a full agonist ( = 0.85). In this lim-
a
lue = 1). Preferred compounds exhibited
and 0.5 in the in vitro D2 functional assay.
a-values between 0.2
ited set of derivatives, there is no clear structure–activity trend in
D2 functional activity.
Dopamine D2 and SRI binding data for each linker series are
shown in Tables 1–3. Table 1 shows results with alkyl linkers at-
tached via oxygen to the lactam SRI fragment. These compounds
consistently show good D2 affinity, and less potent SRI affinity.
Compounds containing sulfonamide D2 moieties (29, 30, 33 and
34) display substantially improved
a1 receptor selectivity com-
pared to the benzodioxolanyl or phenyloxazolone D2 building
blocks.4 It is widely held that antagonist activity at this receptor
O
F
O
S
O
O
S
O
OTs
Br
Br
O
H3C
N
H
O
HN
O
H3C
N
H
O
21
20
19
Figure 3. Structures of D2 building blocks.

2986
Y. Yan et al. / Bioorg. Med. Chem. Lett. 20 (2010) 2983–2986
Table 1
Alkoxy-linked analogs
tional activity of these two compounds differs substantially, as 33
shows D2 partial agonist activity in cells, while 34 is essentially a
full agonist. Compound 33 has a superior receptor profile com-
pared to tetrahydrocarbazoles 1 and 2 and is a candidate for more
detailed in vivo studies, in animal models of antipsychotic and
serotonin reuptake inhibition. Compared with the tetrahydrocar-
bazole compounds 1 and 2,4 selected compounds reported here
(
)
N
N
O
nO
H
N
showed improved a1 selectivity (33 and 34) and comparable dopa-
mine D2 partial agonist activity, with potent SRI affinity.
O
O
HN
In summary, in combination with cyclic linkers, the dihydro-
quinolinone pharmacophore has been shown to display potent
SRI activity and furnished compounds with good D2 receptor affin-
ity using a variety of D2 pharmacophores. The tetrahydrocarbazole
analogs described in the first paper in this series showed variable
D2 affinity.4 Using selected cyclic amine linkers, the dihydroquin-
olinone SRI moiety can furnish potent, dual acting compounds with
attractive in vitro profiles and dopamine D2 partial agonist activity.
A rigid piperidine linker, in the form of a [3.2.1] azabicyclo ring sys-
tem, provides derivatives with superior SRI affinity compared to
alkoxy or simple piperidine linkers. Among the two regioisomeric
azabicycles studied in this work, the 8-azabicyclo isomer was
superior to the 3-isomer in terms of SRI and D2 affinity. Several
compounds in this group furnished dopamine D2 partial agonist
activity in the desired range. Sulfonamide-based D2 moieties pro-
Compd
n
SRI binding (Ki, nM)
D2 binding (Ki, nM)
22
23
24
2
3
4
538
93
94
8.6
8.6
8.3
Standards: D2-haloperidol Ki 3 nM, SRI escitalopram Ki 1 nM. All Ki values represent
averages of at least three independent experiments.
Table 2
Oxy-/methylene-linked 4-piperidine analogs
D2
N
X
N
H
O
vided improved
a1 receptor selectivity, compared to benzodiox-
olanyl and phenyloxazolone derivatives. Compound 33 is
candidate for more detailed in vivo studies to more fully assess
its potential as a dual acting antipsychotic agent.
a
Compd D2
X
SRI (Ki, nM) D2 (Ki, nM) D2 func. (
a
-value)
a
1 (Ki, nM)
25
26
27
28
29
30
19
20
21
19
20
21
C
C
C
O
O
O
1.9
19
8
9
0.38
<0.1
<0.1
0.41
0.21
<0.1
2
NT
NT
2
91
354
20
20
22
30
19
2.5
18
9
Acknowledgement
This manuscript is dedicated to our colleague Jan-Hendrik Rein-
ders who passed away on February 4, 2008.
All Ki values represent averages of at least three independent experiments. a-Values
represent average of at least two independent experiments.
References and notes
1. Tohen, M.; Vieta, E.; Calabrese, J.; Ketter, T. A.; Sachs, G.; Bowden, C.; Mitchell,
P. B.; Centorrino, F.; Risser, R.; Baker, R. W.; Evans, A. R.; Beymer, K.; Dube, S.;
Tollefson, G. D.; Breier, A. Arch. Gen. Psychiatry 2003, 60, 1079.
2. Amsterdam, J. D.; Shults, J. J. Affect. Disord. 2005, 87, 121.
3. Benazzi, F.; Berk, M.; Frye, M. A.; Wang, W.; Barraco, A.; Tohen, M. J. Clin.
Psychiatry 2009, 70, 1424.
4. Rotella, D. P.; McFarlane, G. R.; Greenfield, A.; Grosanu, C.; Robichaud, A. J.;
Feenstra, R. W.; Núñez-Garcia, S.; Reinders, J.-H.; van der Neut, M.; McCreary,
A.; Kruse, C. G.; Sullivan, K.; Pruthi, F.; Lai, M.; Zhang, J.; Kowal, D. M.; Carrick,
T.; Grauer, S. M.; Navarra, R. L.; Graf, R.; Brennan, J.; Marquis, K. L.; Pausch, M.
H. Bioorg. Med. Chem. Lett. 2009, 19, 5552.
5. Chris, L. B.; Bruce, M. P. Curr. Control Trials Cardiovasc. Med. 2002, 3, 7.
6. Melzer, H. Y. Neuropsychopharmacology 1999, 21, 106S.
7. McGavin, J. K.; Goa, K. L. CNS Drugs 2002, 16, 779.
8. Shapiro, D. A.; Renock, S.; Arrington, E.; Chiodo, L. A.; Liu, L. X.; Sibley, D. R.;
Roth, B. L.; Mailman, R. Neuropsychopharmacology 2003, 28, 1400.
9. Stack, G. P. US 5,869,490, 1999.
Table 3
Bridged piperidine analogs
B2:
HN
HN
B1:
L
D2
N
H
O
Compd D2
L
SRI (Ki, nM) D2 (Ki, nM) D2 fn. (
a
-value)
a
1 (Ki, nM)
31
32
33
34
19 B1 0.26
19 B2 1.5
20 B2 1.7
16
6.3
11
7
<0.1
0.4
0.25
0.85
8
46
>3300
1100
21 B2
5
All Ki values represent averages of at least three independent experiments. a-Values
represent average of at least two independent experiments.
10. Galante, R. J. WO2007139998, 2007.
11. Failli, A. A.; Evrard, D.; Li, Y.; Hatzenbuhler, N. T.; Mogish, J.; Gunawan, I. S.;
Nikitenko, A.; Zhou, P., US20070530, 2008.
12. Compound 19 see Ref. 9; compounds 20 and 21: Mewshaw, R. E.; Husbands,
M.; Gildersleeve, E. S.; Webb, M. B.; Shi, X.; Mazandarani, H.; Cockett, M. I.;
Ochalski, R.; Brennan, J. A.; Abou-Gharbia, M.; Marquis, K.; McGaughey, G. B.;
Coupet, J.; Andree, T. H. Bioorg. Med. Chem. Lett. 1998, 8, 295.
13. All target compounds were characterized by NMR, analytical HPLC and MS, and
were at least 97% pure.
could induce sedation, and may lead to cardiovascular issues.5,6
Compounds 33 ( 1 Ki >3300 nM) and 34 ( 1 Ki 1100 nM) were
especially noteworthy in this regard, with substantial separation
between the primary targets and 1 affinity. Interestingly the func-
a
a
a