Non-basic azolotriazinone MCHR1 antagonists for the treatment of obesity: An empirical brain-exposures-driven candidate selection for in vivo efficacy studies

Article abstract of DOI:10.1016/j.bmcl.2015.09.018

Non-basic azolotriazinones were explored using an empirical free brain exposures-driven approach to identify potent MCHR1 antagonists for evaluation in in vivo efficacy studies. An optimized lead from this series, 1j (rMCHR1 K_i = 1.8 nM), demonstrated a 6.9% reduction in weight gain relative to vehicle in a rat model at 30 mg/kg after 4 days of once-daily oral treatment as a glycine prodrug. Despite a promising efficacy profile, an assessment of the biliary toxicity risk of this compound rendered this compound non-progressible.

Full text of DOI:10.1016/j.bmcl.2015.09.018

Bioorganic & Medicinal Chemistry Letters 25 (2015) 4412–4418
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
Non-basic azolotriazinone MCHR1 antagonists for the treatment
of obesity: An empirical brain-exposures-driven candidate selection
for in vivo efficacy studies
a
a
b
b
Pratik Devasthale a, , Wei Wang , James Mignone , Kishore Renduchintala , Sridhar Radhakrishnan ,
⇑
Jayanthi Dhanapal b, Jagannath Selvaraj b, Rajesh Kuppusamy b, Mary Ann Pelleymounter c,
Daniel Longhi c, Ning Huang c, Neil Flynn c, Anthony V. Azzara c, Kenneth Rohrbach c, James Devenny c,
Suzanne Rooney c, Michael Thomas c, Susan Glick c, Helen Godonis c, Susan Harvey c, Mary Jane Cullen c,
Hongwei Zhang d, Christian Caporuscio d, Paul Stetsko c, Mary Grubb c, Christine Huang c, Lisa Zhang c,
Chris Freeden c, Brian J. Murphy c, Jeffrey A. Robl a, William N. Washburn a,
⇑
a Metabolic Diseases Chemistry, Bristol-Myers Squibb Research and Development, Princeton, NJ 08543-5400, United States
b BMS-Biocon Research Center, Biocon, 20th KM, Hosur Road, Electronic City, Bangalore 560 100, India
c Metabolic Diseases Biology, Bristol-Myers Squibb Research and Development, Princeton, NJ 08543-5400, United States
d Pharmaceutical Candidate Optimization, Bristol-Myers Squibb Research and Development, Princeton, NJ 08543-5400, United States
a r t i c l e i n f o
a b s t r a c t
Article history:
Received 13 July 2015
Revised 2 September 2015
Accepted 7 September 2015
Available online 8 September 2015
Non-basic azolotriazinones were explored using an empirical free brain exposures-driven approach to
identify potent MCHR1 antagonists for evaluation in in vivo efficacy studies. An optimized lead from this
series, 1j (rMCHR1 Ki = 1.8 nM), demonstrated a 6.9% reduction in weight gain relative to vehicle in a rat
model at 30 mg/kg after 4 days of once-daily oral treatment as a glycine prodrug. Despite a promising
efficacy profile, an assessment of the biliary toxicity risk of this compound rendered this compound
non-progressible.
Keywords:
Non-basic MCHR1 antagonists
Anti-obesity
Ó 2015 Elsevier Ltd. All rights reserved.
Off-rates
Biliary toxicity
Metabolite toxicity
Melanin-concentrating hormone receptor
1
(MCHR1) is
a
treatment of obesity (Fig. 2). For the azolotriazinones described
in this manuscript, we employed an empirical approach utilizing
pharmacokinetic studies to determine brain concentrations and
brain/plasma (B/P) ratios 8 h after administration of compounds
of interest, dosed as their amino acid prodrug esters 4 (Table 3).
Compounds that met these criteria became candidates for efficacy
studies in our 4-day rat weight loss model.
Schemes 1–3 describe the synthesis of all pyrrolotriazinones (1)
described in this Letter.4 A key intermediate, (Z)-methyl 4-bromo-
1-(((dimethylamino)methylene)amino)-1H-pyrrole-2-carboxylate
(10) was prepared from pyrrole 5 via acylation, bromination, and
amination as shown in Scheme 1.5 Compound 10 was first con-
densed with appropriate anilines under microwave conditions to
assemble the pyrrolotriazinones and then coupled with 4-chloro-
boronic acid using Suzuki–Miyaura conditions6 to yield the desired
targets. The order was reversed in cases where the desired prod-
ucts could not be obtained via the cyclization-Suzuki sequence.
The yields in these cases were particularly low.
G-protein-coupled receptor that is involved in the regulation of
appetite and hence has been actively pursued as a target in
anti-obesity drug discovery.1 Previously disclosed findings from
our MCHR antagonist program obtained with earlier non-basic
thienopyrimidinone anti-obesity agents such as A and B (Fig. 1)
had suggested a general need for free drug concentrations in the
brain to be above the drug’s Ki value at trough.2 In addition, factors
such as receptor occupancy and slower off-rates play a role in
determining efficacy in reducing weight gain.
As part of an expanded effort within our laboratories to identify
novel cores3 and exploit our discovery of potent non-basic
thienopyrimidinones that are devoid of hERG activity, we explored
azolotriazinones 1, 2, and
3 as MCHR1 antagonists for the
⇑
Corresponding authors. Tel.: +1 609 818 4974 (P.D.).
E-mail address: pratik.devasthale@bms.com (P. Devasthale).
http://dx.doi.org/10.1016/j.bmcl.2015.09.018
0960-894X/Ó 2015 Elsevier Ltd. All rights reserved.

P. Devasthale et al. / Bioorg. Med. Chem. Lett. 25 (2015) 4412–4418
4413
Pyrazolotriazinones (2) were prepared via amination of pyra-
zole ester 207 with chloramine in high yield followed by transfor-
mations analogous to the ones described in earlier schemes (see
Scheme 4).
Scheme 5 describes the synthesis of imidazolotriazinones 3a,
3b, and 3c from aminoimidazole ester 25, which is obtained by
condensation of keto acetal 23, ammonium acetate, and methyl
glyoxalate followed by subsequent amination of the imidazole.8
Compounds 3a and 3b were obtained from the appropriately-sub-
stituted amidine 26 under microwave conditions. Compound 3c
was generated via cyclization of the precursor amide 29 using
microwave conditions.
OH
O
O
O
O
O
N
OH
O
N
S
S
N
N
Cl
Cl
A
B
Figure 1. Non-basic thienopyrimidinones.2
Several analogues were obtained using a step-wise sequence of
amide formation followed by cyclization to form the pyrrolotriazi-
none (Scheme 2). In these cases, the Suzuki–Miyaura coupling was
accomplished using bromopyrrole intermediate 10 prior to amide
formation.
Scheme 3 describes a route that reverses the polarity of the cou-
pling partners which allowed for late-stage functionalization via
Suzuki–Miyaura couplings under microwave conditions.
Before embarking on an extensive SAR effort, our first task was
to identify which of the three azolotriazinones (1, 2 and 3) under
consideration was most promising. We evaluated three R groups
(methoxy, pyrrolidinylethoxy, and cyclopropylhydroxyethoxy;
O
O
N
O
R2
N
R2
R2
N
N
N
R1
R1
R1
N
N
N
N
N
N
1
2
3
Pyrrolotriazinones
Pyrazolotriazinones
Imidazolotriazinones
Figure 2. Azolotriazinones.
H
N
H
H
N
b
N
a
COCCl3
COCCl3
Br
5
6
7
NH2
N
H
N
CO Me
e
c
d
2
CO2Me
Br
Br
8
9
R2
O
N
N
NMe2
N
N
N
N
O
g
f
N
Cl
CO2Me
N
R2
OTBDMS
Br
Br
11
10
1a, 1j
O
13, R2 =
OMe
i
h
N
N
NMe2
1c
N
O
j
N
Cl
CO2Me
N
R2
12
1b, 1d, 1k
Cl
k, l, j
N
N
Cl
N
OMe
O
O
OH
1x
Scheme 1. (a) CCl3COCl, Et2O, rt-reflux, 3 h, 91%; (b) Br2, CHCl3, 0 °C, 73%; (c) NaOMe, MeOH, 1 h, 86%; (d) NaH, DMF, 0 °C, 1 h; H2NCl/MTBE, 0 °C, 4 h, 67%; (e) DMF-
dimethylacetal, 60 °C, 3 h, 64%; (f) R2NH2, MeCN/AcOH (2:1), microwave reactor 160 °C, 30 min, 55–75%; (g) 4-chlorophenylboronic acid, Pd(dppf)Cl2, K2CO3, toluene, H2O,
nBu4NOH, 90 °C, overnight, 46% 1a, 11% 1j, 17% 13; (h) TBAF, THF, rt, 3 h, 55%; (i) 4-chlorophenylboronic acid, Pd(dppf)Cl2, K2CO3, toluene, H2O, nBu4NOH, 90 °C, overnight,
44%; (j) R2NH2, MeCN/AcOH (2:1), microwave reactor 160 °C, 30 min, 2% 1b, 2% 1d, 6% 1k, 74% 1x; (k) LiOH, THF–H2O (2:1), 60 °C, overnight, 65%; (l) (R)-1-(4-amino-2-
methoxyphenoxy)propan-2-ol, EDCÁHCl, HOBT, DIPEA, DMF, rt, 20 h, 80%.

4414
P. Devasthale et al. / Bioorg. Med. Chem. Lett. 25 (2015) 4412–4418
N
N
NMe2
CO2Me
N
N
NMe2
CO2H
N
N
NMe2
CO2Me
a
b
R1
R1
Br
10
R1 =
R1 =
R1 =
15a
14a
14b
14c
Cl
15b
15c
Cl
Cl
N
N
N
N
N
NMe2
O
1e, 1f, 1h, 1i, 1l, 1m,
1n, 1p, 1r, 1v, 1w, 1x
d
c
NHR2
15a/15b/15c
X
Y
Cl
16
X, Y = CH or N
Scheme 2. (a) R1B(OH)2, Pd(dppf)Cl2, K2CO3, toluene, H2O, nBu4NOH, 90 °C, overnight, 56% 14a, 23% 14b, 43% 14c; (b) LiOH, THF/H2O, 60 °C, overnight, 65% 15a, 62% 15b, 75%
15c; (c) R2NH2, HOBT, EDCI, DIPEA, DMF, rt, 15 min, 32–86%; (d) MeCN, AcOH, microwave reactor 160 °C, 60 min, 10–75%.
N
N
NMe2
CO2H
N
N
NMe2
N
N
NMe2
O
a
b
CO2Me
NHR2
Br
Br
Br
17
18
10
R2
O
R2
N
N
N
N
N
N
e
d
c
O
1g, 1o, 1q, 1s, 1t, 1u
B
O
Br
O
19
11
Scheme 3. (a) LiOH, THF/H2O, 60 °C, overnight, 70%; (b) R2NH2, HOBT, EDCI, DIPEA, DMF, rt, 15 min, 30–60%; (c) MeCN, AcOH, microwave reactor 160 °C, 60 min, 10–32%; (d)
bis(pinacolato)diboran, KOAc, Pd(dppf)Cl2, DMF, 100 °C, overnight, 45–65%; (e) R1Br, K2CO3, Pd(dppf)Cl2, DMF, microwave reactor 100 °C, 60 min, 6–50%.
Table 1) for each of the three cores, viz. pyrrolotriazinones 1
(X = Y = CH), pyrazolotriazinones 2 (X = CH, Y = N), and imidazolo-
triazinones 3 (X = N, Y = CH). As shown in Table 1, with a small,
basic, and a non-basic R substituent, the pyrrolotriazinones 1 were
consistently superior with respect to in vitro potency and hence
were chosen for further SAR development.
IC50 > 10
pharmacokinetic studies in rat. Encouraged by the finding of 11
brain concentrations of 1i at 8 h (following oral administration of 1i
as its -valine ester prodrug at 10 mg/kg), the compound was
advanced into a 4-day rat efficacy study (Table 4). The reasons
for failure of 1i at 30 mg/kg to produce a statistically significant
reduction in weight in the drug-treated cohort versus the vehicle
group were not clear but we shifted the SAR focus to compounds
containing non-basic substituents.
Our SAR progression sought to enhance in vitro potency and
brain exposure while reducing protein binding. It was anticipated
that free fraction would be enhanced upon incorporation of nitro-
gen atoms within the left-hand aryl R1 group since each nitrogen
atom would lower clog P by approximately 1–1.2 log units
(Table 3). A variety of hydroxy-containing non-basic R2 groups
l
M)10 inhibitory activity and thus was selected for coarse
lM
L
Compound 1b with a 3-methylaminopyrrolidine substituent
was expectedly potent but, not surprisingly, showed significant
hERG activity (hERG flux IC50 = 5.5
lM; electrophysiology assay:
91% at 0.3
l
M).9 In an attempt to modulate ion channel activity,
several hydroxy-substituted cyclic amine substitutions on the
pyridine ring were generated which afforded several potent
MCHR1 antagonists (Table 2). Within this series, 1i exhibited
the most favorable balance between MCHR1 (Ki = 12 nM) and
hERG (hERG flux IC50 > 80 lM; Q-patch electrophysiology assay:
NMe2
CO2Et
c
NH2
N
N
N
H
N
CO2Et
CO2Et
N
N
N
a
b
2a, 2b, 2c
Cl
Cl
Cl
20
21
22
Scheme 4. (a) NaH, DMF, 0 °C, 1 h; H2NCl/MTBE, 0 °C, 4 h, 94%; (b) DMF-dimethylacetal, 60 °C, 3 h, 78%; (c) R2NH2, MeCN/AcOH (1:1), 120 °C, overnight, 6% 2a, 4% 2b, 6% 2c.

P. Devasthale et al. / Bioorg. Med. Chem. Lett. 25 (2015) 4412–4418
4415
Table 2
HO
Cycloalkylaminopyridine-containing pyrrolotriazinones
NH
OH
a
b
Cl
Cl
N
CO2Me
O
23
R
O
N
24
N
N
Cl
N
X
NMe2
R2N
NH2
N
N
26
Compd
X
R
Ki, nM SEM1
hERG flux2
Cl
3a, 3b
c
N
25
CO2Me
Human
Rat
IC50, lM
1b
1d
(see Table 1)
0.5 0.1
0.4 0.2
5.5
d
N
CH
0.4 0.03
15 10
0.3 0.1
14 0.6
3
NHMe
NMe2
1e
1f
CH
CH
N
N
OH
3.3
NMe2
CO2Me
N
N
N
e
Cl
Cl
N
N
CO2H
OH
5.3 1.7
124 49
5.5 1.0
54 19
26% @ 1
>80
l
M2
H
H
28
27
N
N
OH
1g
1h
N
NMe2
N
N
g
H
N
f
Cl
3c
N
O
CH
CH
N
0.8 0.3
12 4.3
1.4 0.8
18 8.2
2
29
O
OH
OH
HO
1i
N
>80 (>10 l
M)3
Scheme 5. (a) NH4OAc, methyl glyoxate, H2O, MeCN, 0 °C—rt, overnight, 34%; (b)
NaH, DMF, 0 °C, 1 h; H2NCl/MTBE, 0 °C, 4 h, 72% (c) 26, MeCN/AcOH (1:1),
microwave reactor 160 °C, 30 min, 9% 3a, 5% 3b; (d) DMF-dimethylacetal, 60 °C,
3 h, 81%; (e) LiOH, THF/H2O, 60 °C, overnight, 70%; (f) 2-(4-aminophenoxy)-1-
cyclopropylethanol, HOBT, EDCI, DIPEA, DMF, rt, 15 min, 37%; (g) MeCN, AcOH,
microwave reactor 160 °C, 60 min, 6%.
1
2
3
n = 3–5.
see note 10.
Q-patch electrophysiology assay data (see note 10).
alcohols) or a L-valine (for secondary alcohols) pro-drug ester (4).
which had previously found to be preferred across multiple
chemotypes were systematically mixed and matched with the
three R1 groups, viz. 4-chlorophenyl, 4-chloro-2-pyridyl, and
4-chloro-2,6-pyrimidyl. Due to the generally poor solubility of
the parent alcohol, coarse pharmacokinetic studies in rats required
administration of each alcohol in Table 3 as a glycine (for tertiary
In general, prodrugs were rapidly cleared from plasma following
oral administration. For example, Cmax for the glycine prodrug of
compound 1j was 5 nM at 1.5 h with an AUC0–8h of 15 nM. Concen-
trations of the prodrug in the brain at 8 h were below the limit of
detection. Plasma Cmax for the parent 1j was 1.57
lM at 4 h with an
AUC0–8h of 8.2 M. For logistical reasons, to allow for a higher
l
throughput of pharmacokinetic (PK) evaluations of compounds,
coarse PK studies were designed to measure plasma concentrations
at 2, 4, and 8 h (data not shown for 2 h and 4 h timepoints), and
brain concentrations at 8 h to provide guidance on Ctrough concen-
trations in plasma and brain. Ctrough concentrations following once-
daily dosing at 20 h were obtained from multi-day efficacy studies.
Hydroxypyrrolidine analog 1k showed brain exposures of
Table 1
Azolotriazinones
R
O
X
Y
N
OMe
Cl
N
N
<1 lM at 8 h. More polar versions of this analog (1l and 1q) had
Compd
X
Y
R
MCHR1 Ki, nM SEM1
Human Rat
lower brain penetration and exposures. Ethyl sulfones 1n and 1r
exhibited particularly poor brain penetration (B/P = 0.1 and 0.03,
respectively). Three compounds (1c, 1j, 1s), each exhibiting brain
concentrations >3 lM at 8 h, were selected for further studies in
the 4-day rat efficacy model (Table 4).
For the pyrrolotriazinones, the CNS free drug exposures above Ki
required to achieve efficacy were much higher than that required
for the thienopyrimidinone standard A despite being full antago-
OMe
OMe
OMe
1a
2a
3a
CH
CH
N
CH
N
21 2.9
86 20
76 32
8
2.0
106 76.8
87 33.0
CH
1b
2b
3b
CH
CH
N
CH
N
0.5 0.1
0.7 0.3
4.4 2.0
0.4 0.2
0.5 0.1
2.6 1.1
N
N
N
O
nists (>90% inhibition at 10 lM). When administered at 3, 10 and
O
30 mg/kg, 1j produced dose-dependent trough free drug brain con-
centrations ranging from >3- to 16-fold over rat Ki and weight
reductions of 3.6–6.9%. While it is known that differences between
brain versus plasma free fractions can be observed, for the pur-
poses of analysis described in this Letter, we have assumed them
to be equivalent.11 Preliminary in-house data for A using rat brain
homogenates suggested a small difference in free fraction in brain
(0.1%) versus plasma (0.3%). However, data from brain homoge-
nates needs to be further corroborated by in vivo microdialysis
experiments.
CH
O
O
1c
2c
CH
CH
N
CH
N
3.5 1.1
13 1.7
26 9.7
2.8 1.1
13 5.9
18 3.5
OH
O
O
OH
OH
3c
CH
In vivo potency of 1j is over an order of magnitude weaker than
A since the trough free drug CNS level multiple over Ki required to
1
n = 3–5.

4416
P. Devasthale et al. / Bioorg. Med. Chem. Lett. 25 (2015) 4412–4418
Table 3
Potency and pharmacokinetic exposures of non-basic pyrrolotriazinones
R2
O
N
N
OMe
R1
N
Compd R1
R2
Ki, nM SEM
Human Rat
Compd
Prodrug esters
Coarse rat PK @ 10 mg/kg
prodrug1
Plasma concn2 @ 8 h,
Brain concn2 @ 8 h,
nM
B/P
ratio
nM
Cl
1j
2.3 1.9
3.5 1.1
4.8 2.4
3.1 0.5
1.8 1.1 4a
2.8 1.1 4b
4.7 0.2 4c
2.3 0.5 4d
Glycine
Valine
903
4551
3742
966
5.0
O
OH
OH
Cl
1c
1248
339
3.0
O
N
Cl
1k
Valine
2.9
0.77
OH
OH
Cl
N
1l
Bis valine
272
210
OH
F
Cl
F
1m
3.4 0.9
3.8 0.7 4e
Glycine
477
2381
0.6
O
O
OH
OH
Cl
1n
2.4 0.9
13 0.9
1.5 0.2 4f
12 3.6 4h
Glycine
Valine
1826
175
0.1
0.2
SO2Et
OH
N
N
Cl
1o
10,200
2042
O
OH
Cl
1p
6.5 1.8 12 12.8 4i
Glycine
7450
1146
0.2
O
N
N
N
1q
1r
Cl
Cl
9.1 2.6
15 7.9
9.6 1.8 4j
8.7 1.5 4k
Valine
4587
664
401
0.1
OH
OH
Glycine
15,595
0.03
O
SO2Et
N
F
F
Cl
1s
6.5 0.2
6.5 0.8 4l
Glycine
5595
3593
1434
0.6
0.5
O
O
OH
N
Glycine @ 5 mg/
kg
Cl
Cl
1t
13 5.7
17 7.0
28 11
15 5.6 4m
15 4.1
2710
ND
OH
OH
N
N
1u
1v
N
N
O
O
OH
Cl
Cl
23 2.4 4n
Glycine
Glycine
3962
125
805
0.03
0.7
SO2Et
F
N
N
F
1w
9.0 2.5
7.9 2.1 4o
1134
N
O
OH
1
2
n = 3–5.
All compounds were dosed orally as
L
-valine (for secondary alcohols) or glycine (for tertiary alcohols) ester prodrugs as suspensions in 0.5% methyl cellulose, 0.1% Tween
80 in water. Exposures refer to parent concentrations.
produce 6.9% weight reduction is 26-fold for 1j versus 1.6-fold for
compound A. Compound 1j was the only pyrrolotriazinone ana-
logue in Table 4 that caused significant weight reduction following
administration at 3, 10 and 30 mg/kg. Compound 1c is even less
effective; when administered at 30 mg/kg only a modest weight
decrease resulted despite the free drug in brain being 16 Â Ki.
The lack of efficacy obtained with doses of 1s at 3 and 10 mg/kg
despite high CNS exposure multiples followed by a sharp 13%
reduction in weight gain relative to vehicle at 30 mg/kg raised con-
cerns of potential toxicity thereby halting any further pursuit. For
these three compounds as well as the basic counterpart 1i, the lack
of correlation between efficacy and concentration of free drug in
the brain above rat Ki suggested that another factor may be playing
a critical role.
A hypothesis has been recently proposed suggesting that MCHR
antagonists exhibiting slow off-rates would be more effective as
anti-obesity agents than those with fast off-rates.3a A qualitative
estimate of relative off-rate characteristics was gauged by
measuring IC50 values in a novel FLIPR assay setup that involves
incubation of test compounds for 120 min followed by

P. Devasthale et al. / Bioorg. Med. Chem. Lett. 25 (2015) 4412–4418
4417
Table 4
In vivo efficacy results with non-basic pyrrolotriazinone MCHR1 antagonists12
Compd Structure
Rat Ki,
nM
%
Prodrug
Dose,
mg/kg
% Weight reduction relative to
vehicle after 4 days
Brain concn @
20 h, nM
Bf2,
nM
Bf/Ki
Free1 esters
3
10
30
3.6*
4.7*
1574
4036
11,679
6.3
16
47
3.5
8.9
26
O
O
N
OH
OH
6.9*
1j
1.8
3
0.4
0.3
Glycine
Valine
N
N
OMe
Cl
N
N
3
10
30
0
0
3418
7976
14,797
10
24
44
3.6
8.6
16
O
2.5*
O
N
1c
OMe
Cl
3
10
30
1.6
4.4
3869
15,974
23,895
63.9
616
60.6
62.5
F
O
F
13*
624
63.7
O
N
OH
1s
6.5
18
60.1
Glycine
Valine
N
N
OMe
Cl
N
3
10
30
1.3
2.2
2.6
1970
10,051
15,108
<3.9
<20
<30
<0.2
<1.1
<1.7
OH
N
O
N
1i
<0.2
N
N
N
Cl
N
N
3
10
30
2.63
3.93
4123
H
OH
12873
O
O
N
7.6*,3
20783
8.4
6
1x
1.4
7.2
0.4
0.7
Valine
OMe
Cl
O
O
O
OH
A
Glycine
10
6.9*
1610
11.3
1.6
S
N
Cl
N
1
2
3
*
Free fraction in rat plasma.
Bf = free brain concentration at 20 h post-dose, assuming % free fraction is equivalent in brain and plasma.
From a 10 day study.
Statistically significant.
measurement of Ca2+ influx upon stimulation with MCH
(Table 5).12 In general, higher T120 IC50 values suggest faster off-
rates. T120 IC50 values for the three pyrrolotriazinones 1c, 1i and
1j are much greater than the value for A. In addition, the progres-
sive decrease in anti-obesity potency tracked with the increase in
T120 IC50 values.
Based on the above efficacy data and overall profiles, 1j was
considered as sufficiently promising to be advanced into further
studies. Metabolite studies with 1j revealed that the livers of rats
Table 5
T120 IC50 values of pyrrolotriazinone MCHR1 antagonists
Compound T120 IC50
nM
,
Weight reduction relative to vehicle after
4 days1
A
16
44
300
5000
6.9% at 10 mg/kg
6.9% at 30 mg/kg
2.5% at 30 mg/kg
No sig. decrease
1j
1c
1i
1
From Table 4.
R
OH
OH
R COOH
OH
OMe
O
O
O
N
O
N
S
S
N
OMe
N
metabolism
metabolism
A
Cl
Cl
(R = CH3)
C (R = H)
30a (R = CH3)
30b
31a (R = CH3)
31b
(R = H)
(R = H)
OH
OH
R
COOH
OH
R
O
O
O
O
N
OMe
1j (R = CH3)
1x
N
OMe
Cl
Cl
N
N
N
N
(R = H)
32a (R = CH3)
32b (R = H)
33a (R = CH3)
33b (R = H)
Figure 3. Potentially toxic metabolites of A, C, 1j, and 1x.

4418
P. Devasthale et al. / Bioorg. Med. Chem. Lett. 25 (2015) 4412–4418
administered the glycine ester pro-drug of 1j at 10 mg/kg con-
tained high levels of diol 32 (10 M) and hydroxyacid 33
(2.4 M) at 20 h post dose (Fig. 3). Prior metabolite investigations
of thienopyrimidinones such as A containing a methylcarbinol
functionality had established that an analogous hydroxylation
and further oxidation of the methyl moiety generated a toxic mix-
ture of diol (30a) and hydroxyacid (31a) that produced biliary
lesions. Moreover, the appearance of lesions was observed after
4 days of treatment of the secondary alcohol counterpart C rather
than 30 days required with tertiary alcohol A.1
References and notes
l
1. (a) Johansson, A.; Löfberg, C. Expert Opin. Ther. Pat. 2015, 25, 193; (b) Högberg,
T.; Frimurerb, T. M.; Sasmal, P. K. Bioorg. Med. Chem Lett. 2012, 22, 6039; (c)
Méndez-Andino, J. L.; Wos, J. A. Drug Discovery Today 2007, 12, 972.
2. Washburn, W. N.; Manfredi, M.; Devasthale, P. V.; Zhao, G.; Ahmad, S.;
Hernandez, A.; Robl, J.; Wang, W.; Mignone, J.; Wang, Z.; Ngu, K.;
Pelleymounter, M. A.; Longhi, D.; Zhao, R.; Wang, B.; Huang, N.; Flynn, N.;
Azzara, A. V.; Barrish, J. C.; Rohrbach, K.; Devenny, J. J.; Rooney, S.; Thomas, M.;
Glick, S.; Godonis, H. E.; Harvey, S. J.; Cullen, M. J.; Zhang, H.; Caporuscio, C.;
Stetsko, P.; Grubb, M.; Maxwell, B. D.; Yang, H.; Apedo, A.; Gemzik, B.; Janovitz,
E. B.; Huang, C.; Zhang, L.; Freeden, C.; Murphy, B. J. J. Med. Chem. 2014, 57,
7509.
l
Before initiating additional studies with 1j, the potential toxic-
ity of the pyrrolotriazinone metabolites was assessed using the
secondary alcohol 1x which was highly metabolized to the corre-
sponding mixture of diol 32b and hydroxyacid 33b. Analogous to
the thienopyrimidinone findings, 100% of the rats administered
the valine ester of 1x at 30 mg/kg for 10 days developed biliary
lesions as well as exhibited a 7.6% reduction in weight gain relative
to vehicle (Table 4). Liver concentrations of the diol (32b) and
3. (a) Devasthale, P.; Wang, W.; Hernandez, A. S.; Moore, F.; Renduchintala, K.;
Sridhar, R.; Pelleymounter, M. A.; Longhi, D.; Huang, N.; Flynn, N.; Azzara, A. V.;
Rohrbach, K.; Devenny, J.; Rooney, S.; Thomas, M.; Glick, S.; Godonis, H.;
Harvey, S.; Cullen, M. J.; Zhang, H.; Caporuscio, C.; Stetsko, P.; Grubb, M.;
Huang, C.; Zhang, L.; Freeden, C.; Li, Y.-X.; Murphy, B. J.; Robl, J. A.; Washburn,
W. N. Bioorg. Med. Chem Lett. 2015, 25, 2793; (b) Preliminary results were
presented at 244th ACS National Meeting, Philadelphia, PA, Aug 19, 2012 Poster
#379.
4. For detailed experimental protocols, see: Devasthale, P.; Washburn, W. N.;
Renduchintala, K. V.; Sridhar, R. WO 2010042682 A1; PCT Int. Appl. 2010.
5. (a) Salvati, M. E.; Illig, C. R.; Wilson, K. J.; Chen, Jinsheng; M., Sanath K.; Wall, M.
J. U.S. Pat. Appl. Publ., US 20040209886 A1, Oct 21, 2004; (b) Hynes, J., Jr.;
Doubleday, W. W.; Dyckman, A. J.; Godfrey, J. D., Jr.; Grosso, J. A.; Kiau, S.;
Leftheris, K. J. Org. Chem. 2004, 69, 1368.
6. Suzuki, A. J. Organomet. Chem. 1999, 576, 147.
7. Wei, F.; Zhao, B.-X.; Huang, B.; Zhang, L.; Sun, C.-H.; Dong, W. L.; Shinc, D. S.;
Miaob, J.-Y. Bioorg. Med. Chem. Lett. 2006, 16, 6342.
8. Chakravarty, P. K.; Fisher, M. H.; Parsons, W. H.; Tyagarajan, S.; Zhou, B. WO
2004094395 A2, Nov 04; PCT Int. Appl., 2004.
9. Huang, C. Q.; Baker, T.; Schwarz, D.; Fan, J.; Heise, C. E.; Zhang, M.; Goodfellow,
V. S.; Markison, S.; Gogas, K. R.; Chen, T.; Wang, X.-C.; Zhu, Y.-F. Bioorg. Med.
Chem. Lett. 2005, 15, 3701.
hydroxyacid (33b) reached 6 lM and 46 lM, respectively. Thus,
as in the case of the thienopyrimidinones, the diol and hydroxyacid
metabolites from pyrrolotriazinone 1j are also implicated in the
observed biliary toxicity.
In conclusion, despite the promising efficacy of 1j, due to the
high potential risk of biliary toxicity, further work on this series
was discontinued. However, the observations on the relationships
between (a) IC50s from a FLIPR assay setup and in vivo efficacy and,
(b) diol and hydroxyacid metabolite formation in the liver and
resultant biliary toxicity, were instructive in guiding our back-up
program toward a candidate nomination for clinical studies.
10. The hERG flux assay was used as an initial screen. The more predictive but
lower throughput Q-patch assay was used for follow-up of compounds of
interest. The Q-patch assay is an automated whole-cell patch clamp
electrophysiology system used to directly measure hERG currents in CHO
cells stably expressing the cloned hERG potassium channel.
11. Kalvass, J. C.; Maurer, T. S.; Pollack, G. M. Drug Metab. Dispos. 2007, 35, 660.
12. For detailed experimental protocols on Ki determination, rat in vivo efficacy
studies, and the FLIPR assay setup, see Ref. 3a. All compounds were dosed
Acknowledgements
We would like to thank Arvind Mathur and Richard Rampulla
for coordinating our collaboration with BMS-Biocon Research
Center, Bangalore, India, and Fang Moore for her help in preparing
this manuscript.
orally once-daily as L-valine (for secondary alcohols) or glycine (for tertiary
alcohols) ester prodrugs as suspensions in 0.5% methyl cellulose, 0.1% Tween
80 in water at the indicated doses.