Tetrahydrocarboline analogs as MCH-1 antagonists

Article abstract of DOI:10.1016/j.bmcl.2010.09.122

A new series of tetrahydrocarbolines with potent MCH-1 antagonist activity were synthesized, using a conformationally constrained design approach towards optimizing pharmacokinetic properties. Two compounds from this series were progressed to a 5-day diet-induced obesity mouse screening model to evaluate their potential as weight loss agents. Both compounds produced a highly significant reduction in weight, which was attributed to their improved pharmacokinetic profile.

Full text of DOI:10.1016/j.bmcl.2010.09.122

Bioorganic & Medicinal Chemistry Letters 20 (2010) 7024–7028
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
Tetrahydrocarboline analogs as MCH-1 antagonists
Alan J. Henderson a, , Dustin Deering a, James F. Grabowski a, Mark Hadden a, Xiaowu Jiang a,
⇑
Yuri Khmelnitsky a, Michele Luche a, Matthew D. Surman a, Sharon Cheetham b, Steven Vickers b,
Jean Viggers b, Peter R. Guzzo a
a AMRI, 26 Corporate Circle, PO Box 15098, Albany, NY 12212-5098, USA
b RenaSci Consultancy Ltd, Biocity, Nottingham, NG1 1GF, UK
a r t i c l e i n f o
a b s t r a c t
Article history:
A new series of tetrahydrocarbolines with potent MCH-1 antagonist activity were synthesized, using a
conformationally constrained design approach towards optimizing pharmacokinetic properties. Two
compounds from this series were progressed to a 5-day diet-induced obesity mouse screening model
to evaluate their potential as weight loss agents. Both compounds produced a highly significant reduction
in weight, which was attributed to their improved pharmacokinetic profile.
Received 20 August 2010
Revised 21 September 2010
Accepted 23 September 2010
Available online 29 September 2010
Ó 2010 Elsevier Ltd. All rights reserved.
Keywords:
Melanin concentrating hormone
MCH
MCH-1
Obesity
Tetrahydrocarboline
Melanin concentrating hormone (MCH) is a peptide originally
identified as a color mediating factor in teleost fish.1 MCH is also
found in mammals as a 19 amino acid cyclic neuropeptide.2 While
mammalian MCH has high homology to salmon MCH, the primary
function is quite different, with the former being primarily local-
ized in the brain.3 Rather than mediating changes in pigment,
mammalian MCH is highly expressed in the lateral hypothalamus
of rodent brains, prompting investigations into the role of the neu-
ropeptide on food intake and body weight regulation.4 Indeed, a
variety of small molecule MCH-1 receptor antagonists have been
shown to effectively reduce body weight in rodent models of obes-
ity.5 As rat and human MCH is identical in structure,6 MCH-1
antagonists represent an exciting target for controlling food intake
and body weight in the overweight and obese population.
was significantly increased (6.0%) when switching to a twice-daily
30 mg/kg dosing regime, suggesting there were opportunities to
improve upon the pharmacokinetic profile of this compound. Thus,
a strategy was employed to improve upon the suboptimal pharma-
cokinetic properties of compound 1, providing compounds with a
profile consistent with once-daily dosing. Our plan was to intro-
duce a conformational constraint between the indazole ring and
basic amine, which would hopefully improve the oral exposure, in-
crease the brain concentration and hence improve upon the effi-
cacy seen for the once-daily dosing of compound 1. It was
believed that such a constrained mimic of the indazole could be
obtained via the c-tetrahydrocarboline scaffold 3 and the b-tetra-
hydrocarboline scaffold 4, both of which were anticipated to have
a basic nitrogen with a similar pKa to that of scaffold 2 (Fig. 1).
Analogs relating to core scaffold 3 were prepared using a
Fisher-indole cyclization as the key step (Scheme 1). 3-Bromo-
phenylhydrazine and 4-N-Boc-piperidone were reacted under
acidic conditions, and the resultant tetrahydrocarboline was re-
protected to provide carbamate 5. Alkylation of the indole was
achieved using sodium hydride and methyl iodide providing 6.
Copper-catalyzed coupling with 4-benzyloxypyridone provided 7,
which could be easily deprotected using methanolic HCl to provide
the desired compound 8. Diverse analogs could be synthesized by
introducing alkyl substituents onto the basic amine (via reductive
alkylation). Additionally, the pyridone functionality could be var-
ied via preparation of suitably substituted pyridones, which them-
selves could be prepared from 4-chloropyridine.8
Previously, we have disclosed the 5-(pyridinin-1-yl)indazole 1
as a potent MCH-1 antagonist that demonstrated significant effi-
cacy in a diet-induced obesity (DIO) mouse screening model.7
N
N
O
N
N
O
1
Compound 1 showed a statistically significant reduction in body
weight (2.8%) at a 60 mg/kg (qd) dose. However, this weight loss
⇑
Analogs of core scaffold 4 were prepared in an analogous man-
ner to that of scaffold 3, except a Fisher-indole cyclization between
Corresponding author.
E-mail address: alan.henderson@amriglobal.com (A.J. Henderson).
0960-894X/$ - see front matter Ó 2010 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2010.09.122

A. J. Henderson et al. / Bioorg. Med. Chem. Lett. 20 (2010) 7024–7028
7025
the potential for tuning the physicochemical properties. Increased
lipophilicity was not, however, a determinant of improved potency
(for example, compound 8 had a c Log P of 3.89 yet was twice as po-
tent as compound 11d with a c Log P of 4.77).
NR
N
X
X
Constraint
N
CH3
N
3
Modification of the pyridone substituent, from a benzyloxy mo-
tif to a directly-linked phenyl motif, also provided potent MCH-1
ligands. Guided by SAR from previously published work,11 where
4-substitution and 2,4-disubstitution was found to be optimal, a
number of analogs were prepared (Table 1). The parent phenyl
compound (11f) showed promising activity, albeit slightly weaker
than that seen for the parent compound in the benzyloxy series (8).
However, introduction of a lipophilic group into the 4-position
significantly improved the MCH-1 binding activity, returning to
potency levels seen within the benzyloxy series (11g–11i). Inter-
estingly, the more polar methoxy (11j) and thiomethyl (11k)
groups were not only tolerated, but delivered two of the most
potent compounds seen within the series. 2,4-Disubstituted phe-
nyl appendages were tolerated (11l–11n), although the potency
was generally slightly weaker than observed for the monosubsti-
tuted counterparts.
N
X
2
NR
CH3
N
4
Figure 1. Strategy for conformationally constraining the indazole and basic amine
of scaffold 2.
NBoc
•HCl
NBoc
a,b
+
Br
NHNH2
O
N
42%
Br
H
5
NBoc
NBoc
SAR around the b-tetrahydrocarbolines (12a–12l) followed a
similar trend to that seen for the c-tetrahydrocarbolines, demon-
strating a level of flexibility around the location of the basic nitro-
c
O
d
N
91%
N
N
51%
Br
O
gen (Table 1). These analogs retained very similar affinity for the
MCH-1 receptor when compared with their c-tetrahydrocarboline
6
7
counterparts. Furthermore, a lack of tolerance towards heterocyclic
phenyl replacements was demonstrated within this series, with
pyridyl analogs (12m and 12n) having weak MCH-1 binding
affinity. This suggests either increased polarity or the presence of
a weakly basic center may not be tolerated within this region of
this particular scaffold. Measurement of the pKa of compound
12a was also undertaken and compared with that of compound 1
to determine whether both scaffolds did indeed contain similarly
basic amines. Compound 12a had a measured pKa of 8.81 which
was in good agreement to that of compound 1 (pKa = 8.63).
A select set of analogs was chosen for further in vitro evalua-
tion, and compared to compound 1. Four compounds (8, 11a, 12a
and 12b) were tested for CYP inhibition, metabolic stability and
solubility (Table 2). All four compounds had an attractive profile,
showing >1000-fold selectivity for MCH-1 over CYP inhibition, very
good aqueous solubility and excellent metabolic stability (with the
exception of 12b in mouse microsomes). Compounds 11a and 12a
were identified as being of particular interest and were selected for
additional in vitro and in vivo studies.
NH
O
e
N
N
95%
O
8
Scheme 1. Reagents and conditions: (a) EtOH, HCl, reflux; (b) Boc2O, Et3N, DMAP,
CH2Cl2, room temperature; (c) NaH, MeI, DMF, room temperature; (d) 4-benzyl-
oxypyridone, CuI, 8-hydroxyquinoline, DMSO, 130 °C; (e) HCl, MeOH, room
temperature.
commercially available 3-bromophenylhydrazine and 4,4-dieth-
oxybutan-1-amine9 was followed by a Pictet–Spengler reaction
using glyoxylic acid10 to generate the tricyclic ring system
(Scheme 2).
The parent compound prepared around the c-tetrahydrocarbo-
line scaffold (8) showed very encouraging MCH-1 potency (Table
1), having similar activity to the indazole (1). SAR development
was conducted around this scaffold with methylation of the basic
nitrogen (11a) retaining excellent potency. Further alkyl substitu-
tion was also tolerated with the hydroxyethylated analog (11b) hav-
ing similar affinity. However, functionalization of this nitrogen
leading to a loss of basicity (as shown by acetamide 11c) caused a
dramatic reduction in potency, highlighting the need for a basic
nitrogen within the structures. Furthermore, functionalization of
the phenyl ring (pertaining to the benzyloxy motif) with lipophilic
4-substituents (11d–11e) led to retention of potency whilst offering
Compound 11a was found to be a highly selective MCH-1
antagonist, having no significant off-target activity (<55% inhibi-
tion @ 1 lM) in a panel of 30 GPCR’s and transporters. Similarly,
compound 12a showed remarkable selectivity in a panel of 88
GPCR’s and transporters, having no significant off-target affinity.
Furthermore, 12a was identified as an MCH-1 antagonist in a
functional assay (IC50 = 14.4 nM).13 Both compounds were then
administered to DIO mice to assess their pharmacokinetic pro-
file. Gratifyingly, the strategy employed to constrain the basic
center in an effort to improve oral exposure proved to be a good
one, as both 11a and 12a had a significantly improved pharma-
cokinetic profile over compound 1. Plasma levels (after 6 h) were
substantially greater than those of 1 at the same dose, and the
brain levels for 12a mirrored this trend (almost three times
higher than those for compound 1). The AUC values for com-
pounds 11a and 12a were also significantly higher than that
for compound 1, further demonstrating improved oral exposure.
Although 12a had significantly higher levels in the brain than 1,
this was attributed to an improvement in oral exposure more
than brain penetration, as the b/p ratios for both compounds
were fairly similar. All three compounds had >90% plasma pro-
tein binding in mouse (Table 3).
•
HCl
NH2
a
Br
Br
NHNH2
94%
N
H
9
b,c
NBoc
14%
N
H
Br
10
Scheme 2. Reagents and conditions: (a) 4,4-diethoxybutan-1-amine, ZnCl2, 180 °C;
(b) (1) glyoxylic acid, KOH, H2O; (2) HCl, H2O, reflux; (c) Boc2O, Et3N, DMAP, CH2Cl2,
room temperature.

7026
A. J. Henderson et al. / Bioorg. Med. Chem. Lett. 20 (2010) 7024–7028
Table 1
SAR of
c
- and b-tetrahydrocarbolines
R
N
R
N
O
O
N
N
N
N
X
X
12
11
Compound
X
R
MCH-1 bindinga,bKi (nM)
Compound
X
R
MCH-1 bindinga,bKi (nM)
1
8
N/A
N/A
H
2.6
O
O
O
O
3.8
4.5
12a
12b
H
4.9
5.2
O
11a
11b
Me
Me
Me
O
–CH2CH2OH
3.6
168
6.7
4.2
15
12c
12d
12e
12f
12g
25
7.6
11
5.6
11
F3C
O
O
11c
11d
11e
11f
–COCH3
Me
H
Cl
Cl
O
H
H
H
F3C
Cl
O
Me
H
Cl
F3C
11g
11h
Me
H
7.8
2.7
12h
12i
H
H
14
F
MeO
MeS
8.4
Cl
F3C
Cl
11i
H
4.2
12j
Me
14
Cl
11j
H
H
H
H
H
2.6
2.4
7.2
7.0
7.6
12k
12l
H
17
MeO
MeS
Cl
Cl
OMe
Me
11k
11l
H
14
MeO
Cl
12m
12n
H
128
708
N
Cl
11m
Me
N
Cl
OMe
Me
F
11n
MeO
a
Displacement of [3H]Compound 1 from MCH-1 expressed in CHO-K1 cells (Kd = 1.42 0.08 nM and Bmax = 13.3 0.7 pmol/mg protein; mean SEM n = 4).12
Values are means of at least two determinations where each determination is within 40% of the mean value shown.
b
In order to assess whether the improved PK of compounds 11a
Compound 12a induced a slightly greater effect, producing a 9.1%
reduction in body weight after dosing at 30 mg/kg (Fig. 3). This reduc-
tion in body weight represented a greater than threefold improve-
ment upon the efficacy seen for compound 1, at half the dose.
In conclusion, a conformational constraint approach was used
to identify tetrahydrocarbolines as potent MCH-1 antagonists.
Select compounds from this class had an excellent in vitro
profile which led to superior pharmacokinetics over the initial
lead compound from our MCH-1 program. These modifications
translated to greatly improved efficacy in a DIO mouse model
of obesity.
and 12a would translate to improved efficacy through once-daily
dosing, both compounds were progressed to a DIO mouse screen-
ing study for analysis of their effects on body weight lowering.
Each compound was dosed over a 5 day period, with the effect
on body weight monitored daily. Compound 11a showed a highly
significant reduction in body weight (7.2%) at
a once-daily
30 mg/kg dose as shown in Figure 2. This compared favorably with
compound 1, which showed a modest reduction in body weight
(2.8%) at double the dose (60 mg/kg) when administered once-
daily.7

A. J. Henderson et al. / Bioorg. Med. Chem. Lett. 20 (2010) 7024–7028
7027
Table 2
Table 3
Additional in vitro data for selected compounds
Comparison of PK data for compounds 1, 11a and 12a in DIO mice
Compound
CYP 3A4aIC50
M)
Half-life (min)
Aqueous solubilityd
(
lM)
PK Profile
1
11a
12a
(l
MLMb
HLMc
Dose (mg/kg)
30 (po)
1950
1377
0.72
14,760
91.4%
30 (po)
5027
—
30 (po)
8026
4012
0.51
41,333
94.1%
[Plasma]6 h (ng/mL)a
[Brain]6 h (ng/g)b
B/P ratioc
1
8
11a
12a
12b
10
68
67
12
24
695
303
462
770
450
492
470
110
>500
93
98.4%e
>1000
482
0.9d
AUC0–6 h (hÃng/mL)e
26,976
98.4%
PPB (mouse)f
63
>1000
a
b
c
d
e
f
Plasma concentration 6 h post dose.
Brain concentration 6 h post dose.
Ratio of brain concentration to plasma concentration at 6 h.
b/p ratio derived from lean mice after a 10 mg/kg oral dose.
Area under curve determined between 0 h and 6 h.
a
1A2, 2B6, 2C9, 2C19, 2D6 and 3A4 isoforms tested. Isoforms not reported have
IC50’s greater than 10 M.
l
b
Mouse liver microsomes.
Human liver microsomes.
Measured in PBS at pH 7.4 after a 24 h equilibration period.
Percentage remaining after 1 h.
c
d
e
Plasma protein binding determined by an ultrafiltration method.
42.0
41.0
40.0
39.0
38.0
***
***
***
***
37.0
Vehicle
***
Compound 11a 30 mg/kg po qd -7.2%
36.0
35.0
34.0
Day -2
Day -1
Day 0
Day 1
Day 2
Day 3
Day 4
Day 5
Day 6
Figure 2. Effect of compound 11a (dosed at 30 mg/kg, po, qd in 1% methylcellulose in water) on the body weight of male C57BL/6J DIO mice. Data are adjusted means (n = 8–
10). SEMs are calculated from the residuals of the statistical model. Data analysed by ANCOVA with body weight on Day 1 as covariate followed by Dunnett’s test for adjusted
data; *p <0.05, **p <0.01, ***p <0.001. Figures in legend refer to % difference from control on Day 6 (i.e., after 5 days dosing).
47.0
46.0
45.0
44.0
***
43.0
***
42.0
***
41.0
***
40.0
39.0
38.0
37.0
Vehicle (po) bid
Compound 12a (30 mg/kg po) qd -9.1%
***
Day -2
Day -1
Day 0
Day 1
Day 2
Day 3
Day 4
Day 5
Day 6
Figure 3. Effect of compound 12a (dosed at 30 mg/kg, po, qd in 1% methylcellulose in water) on the body weight of male C57BL/6J DIO mice. Data are adjusted means (n = 8–
10). SEMs are calculated from the residuals of the statistical model. Data analysed by ANCOVA with body weight on Day 1 as covariate followed by Dunnett’s test for adjusted
data; *p <0.05, **p <0.01, ***p <0.001. Figures in legend refer to % difference from control on Day 6 (i.e., after 5 days dosing).
125, 1660; (c) Qu, D.; Ludwing, D. S.; Gammeltoft, S.; Piper, M.; Pelleymounter,
M. A.; Cullen, M. J.; Mathes, W. F.; Przypek, J.; Kanarek, R.; Maratos-Flier, E.
References and notes
Nature 1996, 380, 243.
1. Kawauchi, H.; Kawazoe, I.; Tsubokawa, M.; Kishida, M.; Baker, B. I. Nature 1983,
5. McBriar, M. D. Curr. Opin. Drug Disc. Dev. 2006, 9, 496.
305, 321.
6. Baker, B. I. Int. Rev. Cytol. 1991, 126, 1.
2. Carpenter, A. J.; Hertzog, D. L. Expert Opin. Ther. Patents 2002, 12, 1639.
7. Surman, M. D.; Freeman, E. F.; Grabowski, J. F.; Hadden, M.; Henderson, A. J.;
3. (a) Skofitsch, G.; Jacobowitz, D. M.; Zamir, N. Brain Res. Bull. 1985, 15, 635; (b)
Jiang, G.; Jiang, X.; Luche, M.; Khmelnitsky, Y.; Vickers, S.; Viggers, J.; Cheetham,
Bittencourt, J. C.; Presse, F.; Arias, C.; Peto, C.; Vaughan, J.; Nahon, J. L.; Vale, W.;
S.; Guzzo, P. R. Bioorg. Med. Chem. Lett. 2010, 20, 7015.
Sawchenko, P. E. J. Comp. Neurol. 1992, 319, 218.
8. Devadas, B.; Walker, J.; Selness, S. R.; Boehm, T. L.; Durley, R. C.; Devraj, R.;
4. (a) Kohler, C.; Haglund, L.; Swanson, S. W. J. Comp. Neurol. 1984, 223, 501; (b)
Hickory, B. S.; Rucker, P. V.; Jerome, K. D.; Madsen, H. M.; Alvira, E.; Promo, M. A.;
Vaugham, H.; Fischer, W. H.; Hoeger, C.; Rivier, J.; Vale, W. Endocrinology 1989,

7028
A. J. Henderson et al. / Bioorg. Med. Chem. Lett. 20 (2010) 7024–7028
Blevis-Bal, R. M.; Marrufo, L. D.; Hitchcock, J., Owen, T.; Niang, W.; Li, X.; Shieh, H.
S.; Sambandam, A.; Liu, S.; Scott, I. L.; McGee, K. F. PCT/US2004/026193,
2004.
obtained from Euroscreen (Batch 1138). Cell membrane homogenates (8.92 lg
protein) were incubated for 60 min at 25 °C with 1.4 nM of the [3H]-labeled
compound in the absence or presence of the test compound in 50 mM Tris–HCl
9. Rinehart, K. L.; Kobayashi, J.; Harbour, G. C.; Gilmore, J.; Mascal, M.; Holt, T. G.;
Shield, L. S.; Lafargue, F. J. Am. Chem. Soc. 1987, 109, 3378.
10. Topai, A.; Rossano, E. C.; Rech, F.; Breccia, P.; Cerbara, I.; Marini, S.; Coletta, M.
PCT/WO2008/148617, 2008.
11. Hadden, M.; Deering, D. M.; Henderson, A. J.; Surman, M. D.; Luche, M.;
Khmelnitsky, Y.; Vickers, S.; Viggers, J.; Cheetham, S.; Guzzo, P. R. Bioorg. Med.
Chem. Lett. 2010, 20, 7020.
12. MCH-1 binding affinity was measured using 4-(3,4,5-tritritiumbenzyloxy)-1-
(1-(2-(pyrrolidin-1-yl)ethyl)-1H-indazol-5-yl)pyridin-2(1H)-one and membranes
prepared from stable CHO-K1 cells expressing the human MCH-1 receptor,
buffer, pH 7.4. Nonspecific binding was determined in the presence of 50 lM 1-
(5-(4-cyanophenyl)bicyclo[3.1.0]hexan-2-yl)-3-(4-fluoro-3-(trifluoromethyl)-
phenyl)-1-(3-(4-methylpiperazin-1-yl)propyl)urea.
13. MCH-1 functional response was measured at Euroscreen assessing Ca2+
mobilization using an Aequorin cell line expressing the MCH-1 recombinant
receptor. Agonist activity was expressed as
a percentage of activity of
reference agonist (MCH) at its EC100 concentration. Antagonist activity was
expressed as percentage inhibition of reference agonist activity at its EC80
concentration.