The discovery of potent, selective, and orally active pyrazoloquinolines as PDE10A inhibitors for the treatment of Schizophrenia

Article abstract of DOI:10.1016/j.bmcl.2011.11.127

High-throughput screening identified a series of pyrazoloquinolines as PDE10A inhibitors. The SAR development led to the discovery of compound 27 as a potent, selective, and orally active PDE10A inhibitor. Compound 27 inhibits MK-801 induced hyperactivity at 3 mg/kg with an ED₅₀ of 4 mg/kg and displays a ～6-fold separation between the ED₅₀ for inhibition of MK-801 induced hyperactivity and hypolocomotion in rats.

Full text of DOI:10.1016/j.bmcl.2011.11.127

Bioorganic & Medicinal Chemistry Letters 22 (2012) 1019–1022
Contents lists available at SciVerse ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
The discovery of potent, selective, and orally active pyrazoloquinolines
as PDE10A inhibitors for the treatment of Schizophrenia
Ginny D. Ho a, , Shu-Wei Yang a, Jennifer Smotryski a, Ana Bercovici a, Terry Nechuta a,
⇑
Elizabeth M. Smith a, William McElroy a, Zheng Tan a, Deen Tulshian a, Brian McKittrick a,
William J. Greenlee a, Alan Hruza b, Li Xiao b, Diane Rindgen c, Deborra Mullins d, Mario Guzzi d,
Xiaoping Zhang d, Carina Bleickardt d, Robert Hodgson d
a Department of Medicinal Chemistry, Merck Research Laboratories, Kenilworth, NJ 07033, USA
b Department of Structural Chemistry, Merck Research Laboratories, Kenilworth, NJ 07033, USA
c Department of Drug Metabolism and Pharmacokinetics, Merck Research Laboratories, Kenilworth, NJ 07033, USA
d Department of Neuroscience, Merck Research Laboratories, Kenilworth, NJ 07033, USA
a r t i c l e i n f o
a b s t r a c t
Article history:
High-throughput screening identified a series of pyrazoloquinolines as PDE10A inhibitors. The SAR devel-
opment led to the discovery of compound 27 as a potent, selective, and orally active PDE10A inhibitor.
Compound 27 inhibits MK-801 induced hyperactivity at 3 mg/kg with an ED50 of 4 mg/kg and displays
a ꢀ6-fold separation between the ED50 for inhibition of MK-801 induced hyperactivity and hypolocomo-
tion in rats.
Received 13 October 2011
Revised 29 November 2011
Accepted 30 November 2011
Available online 9 December 2011
Ó 2011 Elsevier Ltd. All rights reserved.
Keywords:
Phosphodieasterase inhibitors
PDE10A
Schizophrenia
Schizophrenia is a complex thought disorder affecting cognition,
emotion, perception and behavior. It is characterized by ‘positive’
symptoms, including hallucinations, delusions, and disorganized
speech and behavior; ‘negative’ symptoms, including lack of moti-
vation, an inability to express emotion, and apathy, and cognitive
dysfunction. First generation anti-psychotic drugs (APs) are dopa-
mine 2 (D2) receptor antagonists. While effective in treating posi-
tive symptoms, these drugs are ineffective in treating negative
symptoms and cognitive dysfunction and commonly cause extra-
pyramidal syndrome (EPS), a serious side effect characterized by a
constellation of involuntary movements. Second generation APs
target the D2 receptor, as well as other receptors and have a lower
incidence of EPS. However, they are also ineffective in treating neg-
ative symptoms and cognitive dysfunction and are associated with
other severe side effects, including weight gain, diabetes, and QT
prolongation.1–3
D1 receptor agonist, an ideal profile for an AP agent. PDE10A inhib-
itors are under investigation as antipsychotic agents and have been
shown to be efficacious in a variety of preclinical models that may
predict efficacy in the treatment of schizophrenia.15
High-throughput screening identified a series of pyrazoloquino-
lines as PDE10A inhibitors. Pyrazoloquinoline derivatives have
been reported as PDE5 inhibitors previously.16 Compound 1 is a
potent hPDE10A inhibitor with a PDE10A inhibitory activity of
11 nM and selectivity greater than 70-fold over other PDE enzymes
including PDE5A (Fig. 1). It has high ligand efficiency17 (LE = 0.47)
with a calculated lipophilicity (cLogP) of 3.8 and polar surface area
(PSA) of 66. In vitro metabolite identification conducted by incu-
bating compound 1 with rat microsome suggested that O-demeth-
ylation is the major metabolic pathway in the rat. In the mouse and
9
H
Phosphodiesterases (PDEs) regulate intracellular cAMP and/or
cGMP levels and have been identified as potential therapeutic tar-
gets for treating a wide variety of conditions.4–12 PDE10A is a dual
substrate PDE highly expressed in medium spiny neurons of the stri-
atum.13,14 By blocking the degradation of cAMP and cGMP, a PDE10A
inhibitor should mimic the effect of a D2 receptor antagonist and a
8
PDE10A Ki= 11 nM, LE= 0.47
PDE5A Ki > 100,000 nM
N
N1
N
Other PDEs Ki > 770 nM
3
MeO
6
4
N
Rat PK AUC ( PO,10 mg/kg, 0-6 h)19= 3.3 µM.hr
Brain conc: 123 ng/g, B/P ratio= 1.2 (6-h)
Rat PPB (% bound): 84%
3'
N
H
1
4'
MW= 311,cLogP= 3.8, PSA= 66
⇑
Corresponding author.
E-mail addresses: ginny.ho@merck.com, ginny.ho@spcorp.com (G.D. Ho).
Figure 1. Structure, in vitro and PK data of compound 1.
0960-894X/$ - see front matter Ó 2011 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2011.11.127

1020
G. D. Ho et al. / Bioorg. Med. Chem. Lett. 22 (2012) 1019–1022
human microsomes, the major metabolic pathway is hydroxylation
of the piperazine ring. Screening of compound 1 against panels of
receptors and protein kinases indicated that it was active against
serotonin receptors 5HT1b, 5-HT3, and 5-HT7 (54–85% inhibition
Phe729
Ile635
Gln726
Tyr693
at 10 lM) and casein kinase 1 delta (CK1d IC50 = 320 nM). Com-
Zn
pound 1 was evaluated for its anti-psychotic effect in the inhibition
of MK-801-induced hyperactivity assay.18 In this study compound
1 was not active at 30 mg/kg in the rat dosed orally. Therefore the
SAR development was aimed at improving the off-target selectiv-
ity, and in vivo activity of the series.
PO4
Mg
Tyr524
Phe696
His567
Target compounds were prepared according to the procedure
described in Scheme 1. Thiosemicarbazide intermediate 3 was thus
prepared upon treatment of the corresponding aniline with carbon
disulfide and hydrazine. Condensation with ethyl 2-chloroacetoac-
eate gave, following extrusion of sulfur, pyrazole ester 4. Saponifi-
cation and cyclization with POCl3 then yielded the chloride 5. In
order to obtain the title compounds 6, the chloride 5 was allowed
to react with a range of amines with or without solvent at 125 °C or
under microwave condition at 165 °C.20
The PDE inhibitory activity was determined using a scintillation
proximity assay with [3H]cAMP as substrate measuring the hydro-
lysis of cAMP to AMP with recombinant human PDEs 1–11. Selected
compounds were evaluated for PK using the cassette-accelerated
rapid rat screen.19 Anti-psychotic efficacy was evaluated by mea-
suring inhibition of MK-801-induced hyperactivity in the rat.18
The potential for hypolocomotion was evaluated in a spontaneous
locomotor activity test21 to differentiate between efficacy, as de-
fined by reversal of the MK-801 induced behavior, and hypoloco-
motoric effects of the compound.
Figure 2. Co-crystal structure of 1 and PDE10A (PDB ID code 3UUO).
potency, presumably due to the absence of a hydrophobic interac-
tion (Table 1). This clearly demonstrates that these interactions are
important and need to be maintained. The in vitro potency does
not appear to be improved by the addition of a methoxy or chlorine
at the C-6 of the pyrazoloquinoline as in 7 and 10. Replacement of
the methoxy group at the C-6 position with a chlorine (9 and 10) to
address the potential metabolic demethylation produced com-
pounds with similar PDE10A activity. Surprisingly, this modifica-
tion reduced plasma drug exposure of 9 in the rat. The area
under curve (AUC) is 0.14 lM h in the rat PK screen.
Extensive SAR exploration was centered at the C-4 position
(Table 2). Since the 6-OMe-8-Me substitution pattern led to the
best in vitro and PK profile comparing compounds 7–10, this sub-
stitution was maintained for the SAR development to identify the
optimal C-4 substitution. Replacement of piperazine with a piper-
idine (11) or thiomorpholine (14) led to a loss of potency. Morpho-
line (12), homomorpholine (13), and homopiperazine (15) are well
tolerated at the C-4 position. Compound 13 has an excellent
in vitro and PK profile with selectivity greater than 1450-fold over
other PDE enzymes and good plasma drug exposure
The co-crystal structure of compound 1 bound within the cata-
lytic subunit of PDE10A is shown in Figure 2. Several key interac-
tions between compound 1 and the protein were observed. First,
the N-1-H and N-9 of the inhibitor form two hydrogen bonds with
the conserved Gln726. Second, the planar pyrazoloquinoline occu-
pies the hydrophobic clamp defined by Phe729 and Phe696,
(AUC = 11 lM h, B/P = 0.3) in the rat PK screen. It inhibited MK-
achieving favorable p–p stacking through face-to-face interactions
801-induced hyperactivity in the rat at 30 mg/kg dosed orally.
Encouraged by these results, we revisited C-6 position of 13 to
address the potential metabolic demethylation issue. While H,
CH3, F, Cl, Br, OH, and pyrrolidine are tolerated at the C-6 position,
these modifications decreased PDE10A inhibitory activity (Ki = 24–
140 nM) and plasma drug exposure for tested compounds.
with Phe729 and edge-to-face with Phe696. The 8-methyl group
lies underneath Phe729 and displays a hydrophobic interaction
with the protein. The 6-methoxy group is exposed to the solvent.
Although the piperazine points toward the metal ions, they are
separated by a large space filled with a POꢁ4 2 and water molecules,
which allows for possible SAR development in this area.
The subsequent SAR development was aimed at improving the
in vitro potency by incorporating a substituent on piperazine,
piperidine, morpholine, or homopiperazine in 1, 11, 12 and 15.
Small alkyl groups are tolerated at C-30 or C-40 of 1, 11, 12 or 15.
Polar substituents on piperazine, morpholine, piperidine or homo-
piperazine improve PDE10A activity. The biological activities of a
Our initial SAR efforts were directed to probe the importance of
the specific interactions within this chemotype. Introduction of a
methyl group on the N-1 of compound 1 resulted in a significant
loss in potency (PDE10A IC50 = 858 nM), presumably due to disrup-
tion of the H-bond interaction between 1 and Gln726. Similarly, re-
moval of the methyl at the C-8 position (7 and 8) decreased the
Table 1
PDE10A inhibitory activity of C-6 and/or C-8 modification analogs
H
N
H
N
H
N
NHNH2
NH2
b
a
R1
8
N
H
S
N
MeO
MeO
N
MeO
c
O
N
OEt
3
2
R2
4
6
4
N
H
H
N
N
N
N
3'
d
N
H
4'
N
N
MeO
MeO
PDE10A Kia (nM)
NR1R2
Compd
R1
R2
Cl
5
6
7
8
9
10
H
H
Me
H
OMe
H
Cl
585
550
35
Scheme 1. Reagents and conditions: (a) (1) NH4OH, CS2, EtOH, rt; (2) ClCH2CO2H;
(3) NH2NH2, rt, 77% (three steps); (b) CH3C(O)CHClC(O)OEt, EtOH, rt, 64%; (c) (1)
KOH, EtOH, reflux, 5 h; (2) POCl3, (CH2)2Cl2, 57% (two steps); (d) NR1R2, 125 °C or
microwave 165 °C, 10–40%.
Cl
419
a
Values are means of at least two experiments.

G. D. Ho et al. / Bioorg. Med. Chem. Lett. 22 (2012) 1019–1022
1021
Table 2
Table 4
PDE10A inhibitory activity of 4-piperazine modification analogs
PDE10A inhibitory activity of the C-4 secondary amine analogs
H
H
N
N
N
N
N
N
MeO
MeO
4
HN
N
R
( )
3'
nX
Compd
R
PDE10A Kia (nM)
PDE6 Kia (nM)
4'
24
25
26
4-Py-CH2
2-Py-CH2
4-Py-(CH2)2
0.6
6
70
211
1811
554
Compd
X
n
PDE10A Kia (nM)
11
12
13
14
15
CH2
O
O
S
NH
1
1
2
1
2
363
23
3.6
196
15
27
28
29
12
278
697
na
O
52
O
a
Values are means of at least two experiments.
O
( )2
409
30
31
32
130
5
843
representative selection of analogs are shown in Table 3. Com-
pound 16, the (S)-30-Me-40-C(O)NH2 analog of 1 has an improved
PDE10A inhibitory activity (Ki = 3.3 nM) and plasma drug exposure
O
HN
13882
na
O
H3CN
(AUC = 9.5 lM h) in the rat PK screen as compared to 1. However, it
was not active in the in vivo study at 30 mg/kg dosed orally. This
lack of in vivo activity may be due to a low brain level. The brain
drug concentration at an hour after 30 mg/kg oral dosing was
280
O
33
34
55
15
48% @ 3.6
14782
lM
HN
HN
0.027 lg/g. In comparison with 11 or 12, the PDE10A activity
was enhanced by the introduction of a variety of polar groups at
the C-30 or C-40 position (compounds 17–23). In the rat PK study,
no drug concentration of these compounds was detected in the
brain, presumably due to the presence of three or four hydrogen
bond donors in the molecule.
a
Values are means of at least two experiments.
Concurrent SAR development was focused on acyclic amines at
the C-4 position of pyrazoloquinoline. A variety of secondary
amines are tolerated at the C-4 position. The biological activities
of a representative selection of secondary amine analogs are shown
in Table 4. Increasing the linker length between the C-4-N and pyr-
idine (24 vs 26) or tetrahydropyran (27, 28 and 29), or moving the
oxygen of the tetrahydropyran from the C-4 to the C-2 position (27
and 30) led to a loss of potency. However, increasing the linker
length between C-4-N and the piperidine (33 and 34) improved po-
tency. Compound 31 is significantly more potent than 30 and 32,
presumably due to favorable polar interactions between the NH
of the morpholine on 31 and PDE10A. While these modifications
produced potent and selective PDE10A inhibitors, these com-
pounds either exhibited CYP3A4 (24 and 25) or had low plasma
and/or no brain drug exposure in the rat PK screen (31 and 34).
Many of the C-4 secondary amine analogs are potent PDE6
inhibitors. PDE6 activity has known effects on visual disturbances
and is therefore an activity to be avoided.22 A molecular docking
study of 27 and PDE6 suggests a flipped binding mode of the tricy-
clic ring, which places the C-4-NH and C-6-OMe of the pyrazolo-
quinoline next to the conserved Gln773 side chain in PDE6.23
Within this altered mode, the C-4-NH makes a hydrogen bond with
Gln773, but the amine is confined to a relatively tight space. Thus
any bulky substitution at the C-4-NH could decrease PDE6 potency.
Consistent with this binding mode, introduction of a methyl group
on the C-4-NH of 27 or replacing the NH of 27 with an oxygen de-
creased the PDE6 inhibitory activity (35 and 36, Fig. 3). Compound
36 is a potent and selective PDE10A inhibitor with high plasma
Table 3
PDE10A inhibitory activity of ring substitution analogs of 1, 11, and 12
H
N
N
N
drug exposure in the rat PK screen (AUC = 10.6 lM h). In the
N
in vivo study, it inhibited MK-801-induced hyperactivity in the
rat at 30 mg/kg dose orally, but was not active at lower dose
(10 mg/kg).
MeO
3'
R2
X
Compound 27 has a desirable PK profile and good physicochem-
ical properties. It is not a P-glycoprotein substrate in the rat. The
in vitro hepatocyte clearance trend for rat, dog, and monkey for
this compound appears to track with the in vivo exposure. It inhib-
ited MK-801-induced hyperactivity in the rat with an ED50 of 4 mg/
kg. A rat hypolocomotion study was conducted to evaluate the
sedation effect of 27. The ED50 of this compound is 23 mg/kg dem-
onstrating a modest separation between efficacy and hypolocomo-
tion. Compound 27 had undetectable plasma levels (limit of
quantitation = 2 ng/ml) in the monkey, thus the evaluation of the
EPS side effect was hindered as this study is conducted using Cebus
monkeys (Fig. 4).
4'
R1
Compd
X
R1
R2
PDE10A Kia (nM)
Br C6h (ng/g)
16
N
O
C(O)NH2
CH3 (S)
CH2OH
3.3
9.6
<10
<10
(+)ꢁ17
18
C
C
C
C
C
C
NH2
OH
CH2NH2
C(O)NH2
NHC(O)NH2
H
H
H
H
H
1
<10
<10
nd
<10
<10
<10
19
20
21
22
14
0.5
3
27
29
H
CH2OH
(+)ꢁ23
a
Values are means of at least two–three experiments (nd = not determined).

1022
G. D. Ho et al. / Bioorg. Med. Chem. Lett. 22 (2012) 1019–1022
monkey PK for the EPS study will be the subject of future
publications.
H
N
H
N
N
N
N
N
N
MeO
MeO
Acknowledgments
O
The authors would like to thank Drs. John Hunter, Malcolm
MacCoss, Eric Parker and John Piwinski for their support, discus-
sion and suggestions; Dr. Mark Liang and Mr. Tao Meng for prepa-
ration of the intermediate 5.
O
O
36
35
PDE10 Ki= 41 nM
PDE6 Ki= 7383 nM
PDE10 Ki= 3.1 nM
PDE6: 30% inhibition @ 1 µM
Figure 3. Compounds 35 and 36.
References and notes
1. Newcomer, J. W. CNS Drugs 2005, 19, 1.
2. Casey, D. E. Am. J. Med. 2005, 118, 15S.
3. Taylor, D. M. Acta Psychiatr. Scand. 2003, 107, 85.
H
4. Brandon, N. J.; Rotella, D. P. Annu. Rep. Med. Chem. 2007, 42, 3.
5. Hebb, A. L. O.; Robertson, H. A. Curr. Opin. Invest. Drugs 2008, 9, 744.
6. Palmer, M. J.; Bell, A. S.; Fox, D. N. A.; Brown, D. G. Curr. Top. Med. Chem. 2007, 7,
405.
N
N
N
MeO
NH
7. Martin, N.; Reid, P. T. Treat. Respir. Med. 2006, 5, 207.
8. Buchanan Kehler, J.; Ritzen, A.; Greve, D. R. Expert Opin. Ther. Patents 2007, 17,
147.
9. Ueckert, S.; Hedlund, P.; Andersson, K.-E.; Truss, M. C.; Jonas, U.; Stief, C. G. Eur.
Urol. 2006, 50, 1194.
10. Bender, A. T.; Beavo, J. A. Pharmacol. Rev. 2006, 58, 488.
11. Lugnier, C. Pharmacol. Ther. 2006, 109, 366.
12. Jeon, Y. H.; Heo, Y.-S.; Kim, C. M.; Hyun, Y.-L.; Lee, T. G.; Ro, S.; Cho, J. M. Cell.
Mol. Life Sci. 2005, 62, 1198.
13. Fujishige, K.; Kotera, J.; Michibata, H.; Yuasa, K.; Takebayashi, S.; Okumura, K.;
Omori, K. J. Biol. Chem. 1999, 274, 18438.
14. Seeger, T. F.; Bartlett, B.; Coskran, T. M.; Culp, J. S.; James, L. C.; Krull, D. L.;
Lanfear, J.; Ryan, A. M.; Schmidt, C. J.; Strick, C. A.; Varghese, A. H.; Williams, R.
D.; Wylie, P. G.; Menniti, F. S. Brain Res. 2003, 985, 113.
15. Verhoest, P. R.; Chapin, D. S.; Corman, M.; Fonseca, K.; Harms, J. F.; Hou, X.;
Marr, E. S.; Menniti, F. S.; Nelson, F.; O’Connor, R.; Pandit, J.; Proulx-LaFrance,
C.; Schmidt, A. W.; Schmidt, C. J.; Suiciak, J. A.; Liras, S. J. Med. Chem. 2009, 52,
5188.
16. Kumar, V.; Dority, J. A. Jr. PCT Int. Appl. WO 9628159 (1996).
17. Hopkins, A. L.; Groom, C. R.; Alex, A. Drug Discovery Today 2004, 9, 430.
18. Male CD rats were habituated to the locmotor activity (LMA) chambers 24 h
before testing. On test day, they were co-administered MK-801 (0.3 mg/kg, sc)
and the PDE10 inhibitor (30 mg/kg, po) and placed in the testing apparatus.
Inhibition of MK-801 induced activity was measured as a percent of activity in
the PDE10 inhibitor + MK-801group relative to the vehicle + MK-801 group.
19. Korfmacher, W. A.; Cox, K. A.; Ng, K. J.; Veals, J.; Hsieh, Y.; Wainhaus, S.; Broske,
L.; Prelusky, D.; Nomeir, A.; White, R. E. Rapid Commun Mass Spectrom. 2001, 15,
335. The animal was dosed 10 mg/kg orally. The plasma drug exposure was
measured from 0–6 h, and the brain drug concentration was measured after
6 h. The brain/plasma ratio is the ratio of brain and plasma drug concentration
at 6-hour.
O
27
MW= 326, cLogP= 3.5, PSA= 72
PDE10A Ki = 12 nM, LE= 0.45
PDE6 Ki= 278 nM
PDE5A Ki= 17324 nM
Other PDEs Ki > 4793 nM
Rat PK AUC (PO, 10 mg/kg)= 11.8 µM.h
Brain concentration: 1.1 µg/g, B/P ratio= 0.7 (6-h)
Dog PK AUC (PO, 2 mg/kg)= 12.5 µM.h
Monkey PK AUC (PO, 3 mg/kg)= 0 (LOQ= 2 ng/ml)
Hepatocyte clearance (µL/m/M cell): 1.7, 2.9, 0.8, 54 (human, rat, dog, monkey)
Rat PPB (% bound): 93%
Caco2 Efflux ratio= 0.7
Inhibition of MK-801 induced hyperactivity in the rat: ED50= 4 mg/kg
Rat hypolocomotion: ED50= 23 mg/kg
Figure 4. In vitro, in vivo, and PK data of compounds 27.
In summary, crystal structure-aided SAR development has iden-
tified a series of pyrazoloquinolines as potent and selective
PDE10A inhibitors. In general, these compounds have desirable
physicochemical properties (MW <400, cLogP <4, PSA <100). Com-
pound 27 is a potent PDE10A inhibitor with good to excellent
selectivity over other PDE enzymes except for PDE6. Compared
with the lead compound, 27 has an improved selectivity over
CK1d IC50 = 4000 nM). It is efficacious in an animal model believed
to mimic the positive symptoms of schizophrenia, and possesses a
modest separation between the efficacy and hypolocomotion. The
disclosure of further SAR development at the C-4 position of pyraz-
oloquinoline to improve the PDE6 selectivity, in vivo activity and
20. Ho, G. D.; Yang, S-W; Smith, E. M.; Mcelroy, W. T.; Basu, K.; Smotryski, J.; Tan,
Z.; McKittrick, B. A.; Tulshian, D. B. PCT Int. Appl. WO 2010062559 (2010).
21. Hodgson, R. A.; Hyde, L. A.; Guthrie, D. H.; Cohen-Williams, M. E.; Leach, P. T.;
Kazdoba, T. M.; Bleickardt, C. J.; Lu, S. X.; Parker, E. M.; Varty, G. B. Pharmacol.
Biochem. Behav. 2011, 98, 181.
22. Cote, R. H.; Feng, Q.; Valeriani, B. A. Invest. Ophthalmol. 2003, 44, 1524.
23. The PDE6 model is
a homology model constructed using a PDE5 X-ray
structure as the template. The sequence identity between PDE10A and PDE6
proteins at the binding site is 64%.