Synthesis and optimization of N-heterocyclic pyridinones as catechol-O-methyltransferase (COMT) inhibitors

Article abstract of DOI:10.1016/j.bmcl.2016.03.095

A series of N-heterocyclic pyridinone catechol-O-methyltransferase (COMT) inhibitors were synthesized. Physicochemical properties, including ligand lipophilic efficiency (LLE) and c log P, were used to guide compound design and attempt to improve inhibitor pharmacokinetics. Incorporation of heterocyclic central rings provided improvements in physicochemical parameters but did not significantly reduce in vitro or in vivo clearance. Nevertheless, compound 11 was identified as a potent inhibitor with sufficient in vivo exposure to significantly affect the dopamine metabolites homovanillic acid (HVA) and dihydroxyphenylacetic acid (DOPAC), and indicate central COMT inhibition.

Full text of DOI:10.1016/j.bmcl.2016.03.095

Bioorganic & Medicinal Chemistry Letters 26 (2016) 2952–2956
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
Synthesis and optimization of N-heterocyclic pyridinones as
catechol-O-methyltransferase (COMT) inhibitors
b
c
Zhijian Zhao a, Scott T. Harrison a, Jeffrey W. Schubert a, , John M. Sanders , Stacey Polsky-Fisher ,
Nanyan Rena Zhang c, Debra McLoughlin c, Christopher R. Gibson c, Ronald G. Robinson d, Nancy A. Sachs d,
Monika Kandebo d, Lihang Yao d, Sean M. Smith d, Pete H. Hutson d, Scott E. Wolkenberg a,
James C. Barrow a,y
⇑
a Medicinal Chemistry, Merck & Co., Inc., West Point, PA, United States
b Chemistry, Modeling, and Informatics, Merck & Co., Inc., West Point, PA, United States
c Pharmacokinetics, Pharmacodynamics, & Drug Metabolism, Merck & Co., Inc., West Point, PA, United States
d Neuroscience Research, Merck & Co., Inc., West Point, PA, United States
a r t i c l e i n f o
a b s t r a c t
Article history:
A series of N-heterocyclic pyridinone catechol-O-methyltransferase (COMT) inhibitors were synthesized.
Physicochemical properties, including ligand lipophilic efficiency (LLE) and clogP, were used to guide
compound design and attempt to improve inhibitor pharmacokinetics. Incorporation of heterocyclic cen-
tral rings provided improvements in physicochemical parameters but did not significantly reduce in vitro
or in vivo clearance. Nevertheless, compound 11 was identified as a potent inhibitor with sufficient
in vivo exposure to significantly affect the dopamine metabolites homovanillic acid (HVA) and
dihydroxyphenylacetic acid (DOPAC), and indicate central COMT inhibition.
Received 9 February 2016
Revised 24 March 2016
Accepted 25 March 2016
Available online 7 April 2016
Keywords:
Catechol O-methyl transferase
Schizophrenia
Ó 2016 Elsevier Ltd. All rights reserved.
Cognition
Schizophrenia is a chronic, disabling brain disorder estimated to
affect 1% of the population and characterized by positive and neg-
atives symptoms and cognitive impairment.1,2 The latter is pro-
posed to arise from reduced dopamine activity in the prefrontal
cortex.3 Central inhibition of catechol-O-methyltransferase
(COMT), an enzyme responsible for dopamine metabolism, has
been proposed as a mechanism for selectively increasing cortical
dopamine levels4 and modulation of COMT activity has shown
positive effects in rodent models and human clinical trials.5,6
Membrane (MB) and soluble (S) COMT isoforms exist; the former
is predominantly expressed in the brain and the latter in the
periphery.7 Herein, we report efforts to develop a brain-penetrant
MB-COMT inhibitor for the treatment of cognitive symptoms asso-
ciated with schizophrenia. For an extensive review of COMT inhibi-
tion for the treatment of Parkinson’s disease, depression, and other
neurological disorders, please refer to Kiss et. al.8
exhibits potent inhibition of human and rat MB-COMT. Crystallo-
graphic studies with the soluble isoform, S-COMT, identify a criti-
cal interaction between the oxygens of the hydroxypyridinone
pharmacophore and the active-site magnesium (Fig. 1).11 The
N-biphenyl interacts with hydrophobic residues near the active
site and is also partially solvent exposed, leading to speculation
that affinity arises from interaction with the transmembrane por-
tion of MB-COMT that is absent from S-COMT.12 While 1 exhibits
reasonable potency (47 nM), the biphenyl imparts significant
lipophilcity to the polar hydroxypyridinone core. We hypothesized
that suboptimal physicochemical and pharmacokinetic properties
could be improved by identifying more polar replacements of the
hydrophobic biphenyl.
Initial attempts to prepare N-heteroaryl analogs by condensa-
tion of heteroaromatic amines with kojic acid derivatives were
unsuccessful. While useful for preparing N-carbocyclic analogs,9
the electron-deficient nature of the heteroarene presumably
reduces the nucleophilicity of the aryl amine and slows the reac-
tion. Scheme 1 illustrates the alternative route developed, utilizing
benzyl-protected kojic acid derivative 2. Decarboxylation and a
protecting group switch affords pyranone 3, which can be con-
verted to the PMB-protected 3-hydroxy-4-pyridinone 4 by treat-
ment with ammonium hydroxide. Compound 4 participates in
SNAr reactions with halogenated heteroarenes, and subsequent
Harrison et al. explored several heterocyclic bidentate magne-
sium binding motifs and discovered that 3-hydroxy-4-pyrimidi-
none exhibits greater COMT inhibition potency than a similar
5-hydroxy-4-pyrimidinone scaffold.9,10 N-Biphenyl pyridinone 1
⇑
Corresponding author.
y
Present address: Department of Pharmacology and Molecular Sciences, Johns
Hopkins School of Medicine, Baltimore, Maryland, United States.
http://dx.doi.org/10.1016/j.bmcl.2016.03.095
0960-894X/Ó 2016 Elsevier Ltd. All rights reserved.

Z. Zhao et al. / Bioorg. Med. Chem. Lett. 26 (2016) 2952–2956
2953
Mg2+
Mg-binding
pharmacophore
O
N
O
Near-surface
interaction with S-
COMT isoform
MB-COMT IC50 (nM): 47
S-COMT IC50 (nM): 1100
clogP: 2.53
1
Figure 1. N-Biphenyl 3-hydroxy-4-pyridinone COMT inhibitor 1 (blue, PDB 4XUC) bound to S-COMT.
as suggested by S-COMT X-ray structures.8,9,16 While MB-COMT
potency varies greatly (40 nM–5 M), activity against the soluble
isoform varies little (1–5 M), further suggesting that the biaryl
O
O
O
PMBO
)
BnO
PMBO
l
(d
)
a,b,c
(
)
l
N
H
O
CO2H
O
forms additional interaction with MB-COMT.
(g
2
3
4
Because of the improved LLE of compounds 10 and 11, addi-
tional analogs with a 2,4-disubstituted pyridine central ring were
prepared (see Table 2). Bipyridyl analogs (15–17) lose potency
(e,f)
O
O
N
HO
PMBO
O
HO
N
(h,f
)
Table 1
N
N
Ar
N
Potency and LLE data for N-aryl 3-hydroxy-4-pyridinone COMT inhibitors
R
I
15-21
5
6-14
O
HO
Scheme 1. Reagents and conditions: (a) diphenyl ether, 235 °C; (b) HCl, reflux; (c)
PMB-Br, K2CO3, DMF; (d) NH4OH, reflux; (e) ArX, K2CO3, DMSO, 80 °C; (f) 1:1 TFA–
DCM; (g) 2-bromo-4-iodopyridine, K2CO3, DMF; (h) RB(OH)2, Pd(dppf)Cl2ꢁCH2Cl2,
THF, Cs2CO3, H2O.
N
R
a
a
Compound
R
MB-COMT IC50
(nM)
S-COMT IC50
(nM)
LLE
deprotection with TFA affords final compounds 6–14 as summa-
rized in Table 1.13 Alternatively, SNAr with 2-bromo-4-iodopy-
ridine provides intermediate 5. Further derivatization by Suzuki
cross coupling and TFA deprotection provides compounds 15–21
as summarized in Table 2.
1
47
1100
4.79
N
N
6
7
8
706
2054
>5000
2325
4.75
<3.87
4.53
Because improvement in in vivo potency was the goal, ligand
lipophilicity efficiency (LLE), a parameter used to simultaneously
assess potency and physicochemical properties, was used to
identify compounds with improved drug-like properties.14 As
introduced by Leeson et al., LLE = pIC50 (or pKi) ꢀ clogP, and a
value of 5–7 or higher is desirable for oral drugs.15 MB-COMT
potency is used in this calculation, and clogP was calculated using
the AlogP98 method as implemented in CCG (MOE) software.
Both MB- and S-COMT potency proved sensitive to biphenyl
replacements. The 2-phenylpyrimidine 6 loses 15-fold MB-COMT
potency relative to 1. Thiazole analog 7 and fused biaryls (13–14)
lose activity, suggesting a preference for 6-membered central rings.
Comparing 8, 9, and 11 indicates a preference for the 2-pyridi-
none,4-phenyl pyridine substitution pattern.
Consistent with the observation that lipophilic rings in the
biaryl boost potency, fluoro substitution of the central pyridine is
tolerated (10) and a meta relationship of the terminal phenyl is
preferred versus ortho- (compare 9, 11, and 12). Replacing the cen-
tral phenyl with pyridine (10–11) maintains potency, and signifi-
cantly improves LLE relative to 1.
N
N
N
>5000
456
S
N
9
128
55
2584
946
4.54
5.13
5.47
F
10
11
N
40
3020
N
12
>5000
>5000
<3.37
N
N
Interestingly, S-COMT potency generally decreases less than
MB-COMT. Relative to 1, MB-COMT potency shifts 3-10-fold for
8–9, while S-COMT potency for these compounds shifts 2-fold. This
trend continues, with MB-COMT activity varying significantly
across the series, and S-COMT potencies shifting only 1-3-fold.
The preference for hydrophobic groups supports speculation
that the pyridinone N-substituent interacts with the transmem-
brane portion of MB-COMT and is not necessarily solvent-exposed
13
14
962
1268
1667
4.70
N
N
>5000
<4.93
N
a
For assay protocols, see Ref. 18.

2954
Z. Zhao et al. / Bioorg. Med. Chem. Lett. 26 (2016) 2952–2956
Table 2
Potency and physicochemical data for central-pyridine 3-hydroxy-4-pyridinones
O
N
R
HO
N
Compound
R
hMB-COMT IC50 (nM)
40
rMB-COMT IC50 (nM)
15
LLEb
5.47
clogP
11
1.93
N
15
16
826
672
nd
nd
4.88
5.40
1.20
0.78
N
17
18
574
101
nd
15
5.47
4.47
0.78
2.53
N
S
N
19
49.5
28
5.15
2.16
N
N
Cl
20
21
35
84
27
11
4.21
4.50
3.26
2.58
Cl
N
CF3
b
LLE calculated using human MB-COMT IC50 and clogP calcuated using AlogP98 method.
relative to 11, suggesting a preference for hydrophobic or less basic
aryl substituents. Indeed, trifluoromethyl substitution of the termi-
nal pyridine improves potency 10-fold (compare 15 and 21), and
thiazole (18) and pyrazole (19) are tolerated when substituted
with bulky hydrophobic groups. Dichlorophenyl (20) is equipotent
with unsubstituted phenyl (11), but increases compound
lipophilicity. When assessed through lipophilic efficiency, 11
emerges as one of few compounds with substantive improvement
over 1.
Because preclinical biomarker and efficacy studies were
planned using rat models, rat MB-COMT in vitro potency was also
measured, and a consistent trend toward greater potency in rat
was observed. Interestingly, greater improvements in rat versus
human potency were generally observed with lower LLE (compare
18 and 21 with 11 and 19).
Graphical analysis of the above compounds and fifty additional
hydroxypyridinone analogs13 illustrates the relatively narrow
physicochemical property space occupied by potent compounds
(MB-COMT IC50 <200 nM). Excepting 11 (clogP = 1.93), there is an
apparent clogP lower limit of 2, and an apparent trend toward
higher affinity with increasing lipophilicity (Fig. 2a). All com-
pounds prepared have a polar surface area between 40 and
90 Å2, a range that is desirable for brain-penetrant compounds
(Fig. 2c).17 Plotting LLE versus MB-COMT potency illustrates an
upper limit of LLE = 5.5 and further demonstrates that small
potency improvements relative to 11 are generally accompanied
by increasing lipophilicity.
Suitability for in vivo pharmacokinetic and pharmacodynamic
studies was further evaluated by measuring in vitro stability in
human hepatocytes. In vitro intrinsic clearance for this series spans
a broad range and is not dependent on clogP (Fig. 2d). This is con-
sistent with metabolite identification studies indicating pyridinone
glucuronidation, located distal to the site of structural changes in
these analogs, as the major route of metabolism.
Intrinsic clearance values for inhibitors with an IC50 <100 nM are
moderate to high and highlight the difficulty of developing high-
affinity compounds with desirable pharmacokinetic properties
(Table 3). While replacing the biphenyl with phenylpyridine
(compare 1 and 11) lowers clogP and provides small improvements
in human and rat intrinsic clearance, it is accompanied by an
increase in plasma clearance. This suggests that introducing polarity
into the central ring does not significantly lower the rate of pyridi-
none glucuronidation. Further modifications of the terminal aro-
matic ring maintain potency (20), but increase lipophilcity and
intrinsic clearance. Following intravenous dosing in rat (0.5 mg/kg
in DMSO), 11 has an observed plasma clearance = 94 mL/min/kg,
volume of distribution = 1.9 L/kg, and half-life = 0.4 h.
As previously reported, central COMT inhibition alters the con-
centrations of dopamine metabolites, and leads to an increased
concentration of dopamine dihydroxyphenylacetic acid (DOPAC)
and a decreased concentration homovanillic acid (HVA).18 Levels
of HVA and DOPAC in rat cerebrospinal fluid were measured one
hour after intraperitoneal dosing of 11 (Fig. 3) in 90% PEG 400. A
dose-dependent increase in DOPAC concentration and decrease in

Z. Zhao et al. / Bioorg. Med. Chem. Lett. 26 (2016) 2952–2956
2955
Figure 2. Physical properties of hydroxypyridinone COMT inhibitors plotted against MB-COMT inhibition and human intrinsic clearance; compound 11 denoted by black star.
Colored by human in vitro hepatocyte Clint (mL/min/kg): green (60–150), yellow (151–250), red (250–360). cIntrinsic clearance value derived from in vitro human hepatocyte
studies.
Table 3
Human and rat hepatocyte derived metabolic stability, and rat plasma clearance values
O
HO
N
R
Compound
R
Human Clint
Rat Clint
583
Rat Clplasma
1
320
74
58
N
N
11
448
94
20
292
Nd
200
Cl
Cl
All values reported in mL/min/kg. Plasma clearance measured following 0.5 mg/kg dose in DMSO.

2956
Z. Zhao et al. / Bioorg. Med. Chem. Lett. 26 (2016) 2952–2956
Figure 3. CSF levels of DOPAC and HVA one hour after IP dosing of 11. (*p <0.05 compared to vehicle, ANOVA.)
Table 4
References and notes
Plasma and CSF concentration one hour after IP dosing of 11
1. Saha, S.; Chant, D.; Welham, J.; McGrath, J. PLoS Med. 2005, 2, 413.
2. Stratta, P.; Daneluzzo, E.; Bustini, M.; Prosperini, P.; Rossi, A. Schizophr. Res.
2000, 44, 57.
3. Goldman-Rakic, P. S.; Castner, S. A.; Svensson, T. H.; Siever, L. J.; Williams, G. V.
Psychopharmacology 2004, 174, 3.
Dose (mg/kg)
Plasma concentration (nM)
CSF concentration (nM)
30
100
70
3600
BLQ
410
BLQ = below limit of quantitation.
4. Chen, J.; Song, J.; Yuan, P.; Tian, Q.; Ji, Y.; Ren-Patterson, R.; Liu, G.; Sei, Y.;
Weinberger, D. R. J. Biol. Chem. 2011, 286, 34752.
5. Tunbridge, E. M.; Bannerman, D. M.; Sharp, T.; Harrison, P. J. J. Neurosci. 2004,
24, 5331.
6. Tsapakis, E. M.; Giakoumaki, S.; Roussos, P.; Chrysoulaki, A.; Kopsahili, I.;
Koudas, V.; Bitsios, P. Schizophr. Res. 2010, 117, 493.
7. Reenila, I.; Mannisto, P. T. Med. Hypotheses 2001, 57, 628.
8. Kiss, L. E.; Soares-da Silva, P. J. Med. Chem. 2014, 57, 8692.
9. Harrison, S. T.; Poslusney, M. S.; Mulhearn, J. J.; Zhao, Z.; Kett, N. R.; Schubert, J.
W.; Melamed, J. Y.; Allison, T. J.; Patel, S. B.; Sanders, J. M.; Sharma, S.; Smith, R.
F.; Hall, D. L.; Robinson, R. G.; Sachs, N. A.; Hutson, P. H.; Wolkenberg, S. E.;
Barrow, J. C. ACS Med. Chem. Lett. 2015, 6, 318.
HVA concentration was observed, with changes reaching statistical
significance at the highest dose. Plasma and CSF concentrations of
11 were measured following 30 mg/kg and 100 mg/kg IP doses
and, consistent with HVA and DOPAC changes, higher levels were
observed at the upper dose (Table 4). The highest dose leads to a
statistically significant increase in DOPAC and decrease in HVA that
coincides with higher plasma and CSF exposure and suggests inhi-
bition of central MB-COMT in vivo.
In conclusion, 11 was identified as an N-heterocyclic biaryl
pyridinone inhibitor with sufficient potency and pharmacokinetics
to have a significant effect on the dopamine metabolites HVA and
DOPAC and indicate target engagement of central COMT.
Heterocyclic replacements to the biphenyl of 1 retained potency
while lowering clogP 0.6 units and increasing LLE 1.5 units. While
attempts to optimize physicochemical properties were relatively
successful, these changes did not translate into pharmacokinetic
improvements. For potent inhibitors, modification of molecular
properties via alterations to the pyridinone N-substituent did not
affect in vitro intrinsic clearance or in vivo clearance. Additional
efforts to optimize in vivo inhibition of MB-COMT at Merck and
Cerocor, Inc., will be reported in future communications.19
10. Borchard, R. T. J. Med. Chem. 1973, 16, 581.
11. Vidgren, J.; Svensson, L. A.; Liljas, A. Nature 1994, 368, 354.
12. Bai, H. W.; Shim, J. Y.; Yu, J.; Zhu, B. T. Chem. Res. Toxicol. 2007, 20, 1409.
13. Wolkenberg, S.; Barrow, J.C.; Poslusney, M.S.; Harrison, S.T.; Trotter, B.W.;
Mulhearn, J.; Nanda, K.K.; Manley, P.J.; Zhao, Z.; Schubert, J.W.; Kett, N.;
Zartman, A. World patent application WO2011/109254, 2011.
14. Hopkins, A. L.; Keseru, G. M.; Leeson, P. M.; Rees, D. C.; Reynolds, C. H. Nat. Rev.
Drug Disc. 2014, 13, 105.
15. Leeson, P. D.; Springthorpe, B. Nat. Rev. Drug Disc. 2007, 6, 881.
16. Ehler, A.; Benz, J.; Schlatter, D.; Rudolph, M. G. Acta Crystallogr. Sect. D: Biol.
Crystallogr. 2014, 70, 2163.
17. Wager, T. T.; Hou, X.; Verhoest, P. R.; Villalobos, A. ACS Chem. Neurosci. 2010, 6,
435.
18. Robinson, R. G.; Smith, S. M.; Wolkenberg, S. E.; Kandebo, M.; Yao, L. H.; Gibson,
C. R.; Harrison, S. T.; Polsky-Fisher, S.; Barrow, J. C.; Manley, P. J.; Mulhearn, J. J.;
Nanda, K. K.; Schubert, J. W.; Trotter, B. W.; Zhao, Z. J.; Sanders, J. M.; Smith, R.
F.; McLoughlin, D.; Sharma, S.; Hall, D. L.; Walker, T. L.; Kershner, J. L.; Bhandari,
N.; Hutson, P. H.; Sachs, N. A. ACS Chem. Neurosci. 2012, 3, 129.
19. Cerocor Acquires Rights to Merck COMT Inhibitors. http://www.cerecor.com/
news-publications/news-3-20-2013.php.
Acknowledgements
We thank Ray McClain, Anna Dudkina, April Cox, and Semhal
Berhane for purifying novel analogues.