Design, synthesis and evaluation of MCH receptor 1 antagonists - Part I: Optimization of HTS hits towards an in vivo efficacious tool compound BI 414

Article abstract of DOI:10.1016/j.bmcl.2015.05.077

Abstract Despite recent approvals of anti-obesity drugs there is still a high therapeutic need for alternative options with higher efficacy in humans. As part of our MCH-R1 antagonist program for the treatment of obesity, a series of biphenylacetamide HTS

Full text of DOI:10.1016/j.bmcl.2015.05.077

Bioorganic & Medicinal Chemistry Letters 25 (2015) 3264–3269
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
Design, synthesis and evaluation of MCH receptor 1
antagonists—Part I: Optimization of HTS hits towards
an in vivo efficacious tool compound BI 414
a
a
a
a
Stephan G. Müller a, , Armin Heckel , Jörg T. Kley , Thorsten Lehmann , Philipp Lustenberger ,
Thorsten Oost a, Gerald J. Roth a, Klaus Rudolf a, Kirsten Arndt b, Martin Lenter b, Ralf R. H. Lotz c,
Gerd-Michael Maier c, Michael Markert c, Marcus Schindler b, , Dirk Stenkamp a
⇑
a Boehringer Ingelheim Pharma GmbH & Co. KG, Department of Medicinal Chemistry, Birkendorfer Str. 65, 88397 Biberach an der Riss, Germany
b Boehringer Ingelheim Pharma GmbH & Co. KG, Department of Cardiometabolic Research, Birkendorfer Str. 65, 88397 Biberach an der Riss, Germany
c Boehringer Ingelheim Pharma GmbH & Co. KG, Department of Drug Discovery Support Research Germany, Birkendorfer Str. 65, 88397 Biberach an der Riss, Germany
a r t i c l e i n f o
a b s t r a c t
Article history:
Available online 6 June 2015
Despite recent approvals of anti-obesity drugs there is still a high therapeutic need for alternative options
with higher efficacy in humans. As part of our MCH-R1 antagonist program for the treatment of obesity, a
series of biphenylacetamide HTS hits was evaluated. Several issues of the initial lead structures had to be
resolved, such as potency, selectivity over related GPCRs and P-gp efflux limiting brain exposure in this
series. We could demonstrate that all parameters can be significantly improved by structural modifica-
tions resulting in BI 414 as a potent and orally available MCH-R1 antagonist tool compound with
acceptable in vivo efficacy in an animal model of obesity.
Keywords:
Melanin-concentrating hormone (MCH)
MCH-R1 antagonists
Alkynes
Ó 2015 Elsevier Ltd. All rights reserved.
Obesity is a major risk factor for many serious diseases of the
metabolic syndrome complex such as type 2 diabetes, dyslipi-
demia, coronary heart disease and stroke.1 These diseases are a
major burden for patients and health care systems worldwide.
Consequently, both in academia and pharmaceutical industry
research activities focus on understanding the underlying mecha-
nism of the metabolic syndrome in order to develop strategies
for efficacious treatments. As a potential target melanin-concen-
trating hormone (MCH) has been in the focus of obesity research
in the recent years.2 MCH is expressed in brain regions relevant
for food intake and energy homeostasis and is the natural ligand
for both the MCH-R1 and MCH-R2 receptor. Activation of the
MCH-R1 with MCH results in an increase in food intake and body
weight in rodents, therefore antagonizing this effect is considered
as a new concept for the treatment of obesity. As MCH-R2 is
expressed in humans only, its overall physiological role still needs
to be clarified. Despite many efforts in the pharmaceutical indus-
try, a clinical proof of concept has not been achieved yet. The
few compounds advanced to phase 1 clinical stage so far had to
be discontinued due to safety concerns.3–6 To finally assess the
potential of MCH-1R antagonists as a treatment option for obesity
in humans the need for potent and safe compounds is still high.
An in house high throughput screening (HTS) campaign
resulted in the identification of 1 as a moderately potent MCH-
R1 antagonist (IC50 = 251 nM).
A significant improvement in
potency could be achieved by inverting the central amide moiety
(2, IC50 = 15 nM). In order to avoid anilines as potential mutagenic
metabolites replacements by non-aromatic amines were investi-
gated. The use of phenethylamines in combination with a shorter
benzylic left hand amine moiety yielded 3 as a potent MCH-R1
antagonist (IC50 = 2 nM). This series was accepted as a lead series
and optimization efforts were initiated (see Fig. 1).
The synthesis of these compounds started with commercially
available phenethylamine 4. Amide coupling followed by hydroly-
sis gave acid 6, which was further reduced to benzylic alcohol 7.
Activation of the alcohol as mesylate followed by substitution with
primary or secondary amines yielded the desired derivatives (see
Scheme 1).
⇑
Corresponding author. Tel.: +49 7351 540; fax: +49 7351 54 5181.
Representative structure–activity relationships (SAR) in this
series (amine-, left-hand aryl- and right-hand aryl moiety) are
summarized in Table 1. The amine part can be modified in broad
fashion (9, 10) and additional functional groups (11, 12) are
E-mail
address:
stephan_georg.mueller@boehringer-ingelheim.com
(S.G. Müller).
Present address: AstraZeneca Mölndal, Innovative Medicines & Early Develop-
ment, Cardiovascular
Mölndal, Sweden.
& Metabolic Diseases iMed, Pepparedsleden 1, SE-43183
http://dx.doi.org/10.1016/j.bmcl.2015.05.077
0960-894X/Ó 2015 Elsevier Ltd. All rights reserved.

S. G. Müller et al. / Bioorg. Med. Chem. Lett. 25 (2015) 3264–3269
3265
O
tolerated. Comparable potencies could be obtained with sub-
stituents in the left-hand aryl ring (13, 14); polar moieties (15)
led to reduced potency. For each right-hand aryl ring small lipophi-
lic substituents proved to be advantageous, with compounds 17
and 18 being the most potent in this series.
O
N
N
O
N
N
O
O
A more in-depth characterization of 3 revealed off-target affini-
ties to both a serotonin receptor (5-HT2A) and a muscarinic recep-
tor (M1) and a potent inhibition of the cytochrome P450 enzyme
CYP 2D6 (see Table 2). Further structure–activity relationship stud-
ies indicate that the amine part has a strong influence on the selec-
tivity profile. The introduction of an amide (22, 23) or urea moiety
(26) was beneficial for selectivity. In addition, N-methylation of the
central amide further reduced 5-HT2A affinity compared to the
secondary amide (24 vs 23). The CYP 2D6 interaction potential
was also reduced by the introduction of the polar moieties (e.g.,
24). However, despite the increased polarity these compounds
exhibited only low to moderate metabolic stability in human and
rat liver microsomes.
1
2
O
Cl
H
N
N
O
A rat pharmacokinetic (PK) investigation of 3 as a prototypical
compound of this series revealed a low brain-to-plasma ratio
which could be significantly increased with a P-glycoprotein (P-
3
Figure 1. HTS hit series and lead series.
Table 1
Representative SAR for lead series
Y
3
R
N
2
R
N
X
1
R
2
A
O
3
R1R2N
A
R3
X
Y
MCH-R1a
IC50 (nM)
3
9
Pyrrolidin-1-yl
Amino
Piperidin-1-yl
2-Hydroxyethylamin-1-yl
(Dimethylaminocarbonylmethyl)ethyl amin-1-yl
Pyrrolidin-1-yl
Pyrrolidin-1-yl
Pyrrolidin-1-yl
Pyrrolidin-1-yl
Pyrrolidin-1-yl
Pyrrolidin-1-yl
H
H
H
H
H
H
H
H
H
H
H
H
H
Me
H
H
H
H
H
H
H
H
H
H
H
H
H
4-Cl
4-Cl
4-Cl
4-Cl
4-Cl
4-Cl
4-Cl
4-Cl
4-Cl
4-Cl
4-Cl
3-Cl
4-F
2
68
2
2
8
6
2
470
8
1
10
11
12
13
14
15
16
17
18
19
20
21
3-OMe
2-Br
3-CONH2
H
H
H
H
H
H
H
2-F
3-F
H
H
H
1
29
3
Pyrrolidin-1-yl
Pyrrolidin-1-yl
Pyrrolidin-1-yl
2,4-Cl2
10
a
For the description of biological methods, see Ref. 7.
Cl
Cl
H
N
H
N
a,b
H
O
H
H
O
+
O
O
O
O
O
5
6
4
Cl
Cl
H
N
H
N
c
R2
1 N
d
O
HO
O
R
8
7
Scheme 1. Reagents and conditions: (a) SOCl2, cat. DMF, 60 °C; (b) 2 M NaOH, EtOH rt; (c) CDI, THF, NaBH4, H2O, rt; (d) NEt3, CH2Cl2, MeSO2Cl, 1 h 40 °C then R1R2NH, DMF, rt.

3266
S. G. Müller et al. / Bioorg. Med. Chem. Lett. 25 (2015) 3264–3269
Table 2
Optimization of lead series towards improved selectivity and reduced CYP 2D6 interaction potential
Y
R3
N
R2
X
1 N
O
R
R1R2N
R3
X
Y
MCH-R1a
IC50 (nM)
M1
IC50 (nM)
5HT2A
IC50 (nM)
CYP 2D6
HLM/RLM
t1/2 (min)
IC50
(lM)
3
Pyrrolidin-1-yl
4-Acetylpiperazin-1-yl
4-Acetylaminopiperidin-1-yl
4-Acetylaminopiperidin-1-yl
(4-Acetyl)methylaminopiperidin-1-yl
Ethylaminocarbonyl-methylamin-1-yl
H
H
H
Me
H
H
H
H
H
H
2-F
4-Cl
4-Cl
4-F
4-F
4-F
2
6
3
11
5
84
877
8620
>10,000
2520
>10,000
107
2330
360
>10,000
528
8150
0.5
NT
6.5
15
2.4
>50
NT
NT
>45/22
9/13
12/7
NT/21
22
23
24
25
26
H
4-Cl
36
a
For the description of biological methods, see Ref. 7; NT: not tested; HLM: human liver microsomes; RLM: rat liver microsomes.
O
indicated in 27). The resulting structure–activity relationships are
summarized in Table 3. As both the secondary and tertiary amide
display comparable potency, we designed alkene analogs 29 and
30 to reflect both possible amide conformations. In addition, a
switch to carbon linkers would avoid anilines as potential muta-
genic metabolites. Unfortunately, a pronounced loss in potency
was seen for both modifications, which also held true for the com-
pound with a saturated spacer (31). Surprisingly, the potency could
be regained with an alkyne moiety (32). The hydrogen bond accep-
tor (A@N) seems to be of importance as a 7-fold loss in potency
was observed for the corresponding phenyl derivative 33
(A@CH). The addition of a second acceptor as in 34 resulted in a
reduction of potency. The encouraging results observed for the
alkyne spacer were the basis for further optimization efforts.
The synthesis of the required alkyne derivatives started with
commercially available 2,5 dibromo-pyridine 35. Sonogashira cou-
pling with trimethylsilylacetylene followed by a Suzuki coupling
with substituted aryl-boronic acids and deprotection gave alkyne
intermediates 37. 4-Iodophenol was alkylated to furnish 38 which
was used in a second Sonogashira reaction with 37 to yield 39.
Activation of the alcohol as mesylate (40) followed by a substitu-
tion with primary or secondary amines resulted in the desired
alkyne derivatives 41 (Scheme 2).
The structure–activity relationships for the alkyne derivatives
are summarized in Table 4. Systematic modifications based on
the structure–activity relationships observed in the amide series
(pyrrolidine/4-chloro as in 3 instead of diethylamine/4-methoxy
as in 2) led to the identification of compound 43 as a potent
MCH-R1 antagonist. The SAR of the amine moiety parallels that
of the amide series; broad variations are possible including larger
ring systems (45), introduction of polarity (46, 47) and a second
basic center (48). The attachment of the side chain via carbon
(49) or nitrogen (50) is allowed, even a shorter linker (51) resulted
in compounds of comparable potency. The introduction of nitrogen
in the left hand aromatic system is allowed (52, 53), whereas a
pyridazine moiety led to an observed reduction in potency (54).
Small substituents both in position 2 and 3 were tolerated (55,
56). In the right hand aromatic system a different position for
the hydrogen bond acceptor was not beneficial (57). On the other
hand, increasing polarity by adding ring nitrogen atoms had only
a small influence on potency (58, 59). Substitution with small, lipo-
philic groups was tolerated, with fluorine being the most potent
(60).
N
N
O
O
2
O
X
Y
N
N
O
27
Figure 2. Spacer modifications.
gp) inhibitor (data not shown), thus, indicating 3 to be a substrate
of the P-gp transporter. We felt that an optimization of this series
would require substantial efforts to overcome all the described lia-
bilities. Therefore we started a re-investigation of 2 in order to
identify a new lead series (see Fig. 2).
Hypothesizing that the polar amide linker acts as a recognition
element for P-gp (2 is also a P-gp substrate), we decided to intro-
duce less-polar linker moieties and re-balance the net polarity by
incorporation of heteroatoms in the right hand aryl moiety (as
Table 3
Spacer modifications
O
B
X
Y
A
N
O
–X–Y–
A
B
MCH-R1a
IC50 (nM)
2
–NMe–CO–
–NH–CO–
–CH@CH– (cis)
–CH@CH– (trans)
–CH2–CH2–
–C„C–
CH
CH
N
N
N
N
CH
N
CH
CH
CH
CH
CH
CH
CH
N
15
33
28
29
30
31
32
33
34
700
420
360
35
247
149
In-depth characterization of 43 as a key representative of the
alkyne series revealed similar off-target affinities to the 5-HT2A
receptor and the M1 receptor as seen for the amide series
(Table 5). With modifications in the amine moiety an increase in
–C„C–
–C„C–
a
For the description of biological methods, see Ref. 7.

S. G. Müller et al. / Bioorg. Med. Chem. Lett. 25 (2015) 3264–3269
3267
Y
Br
Br
b,c
a
Br
N
N
N
Si
37
36
35
Y
I
N
e
d
I
O
H
O
H
O
H
O
39
O
38
Y
Y
g
f
N
N
2
R
N
1
R
O
O
40
O
S
O
O
41
Scheme 2. Reagents and conditions: (a) trimethylsilylacetylene, Pd(PPh3)2Cl2, CuI, NEt3, THF, 0 °C?rt; (b) (Y)ArB(OH)2, Pd(PPh3)4, 2 M aq Na2CO3, dioxane, 90 °C; (c)
TBAF*3H2O, THF, rt; (d) 2-chloroethanole, K2CO3, DMF, rt; (e) 37, Pd(dppf)Cl2*CH2Cl2, CuI, piperidine, THF, rt; (f) MeSO2Cl, NEt3, THF, 0 °C?rt; (g) R1R2NH, DMF, 60 °C.
Table 4
Representative SAR of alkyne series
Y
X
Q
E
G
W 2
2
3
R
N
1
R
D
A
B
R1R2N
A
B
D
W
E
G
Q
X
Y
MCH-R1a
IC50 (nM)
32
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Diethylamin-1-yl
Pyrrolidin-1-yl
Pyrrolidin-1-yl
O
O
O
O
O
O
O
O
CH2
NH
—
O
O
O
O
O
O
O
CH
CH
CH
CH
CH
CH
CH
CH
CH
CH
CH
N
CH
CH
CH
CH
CH
CH
CH
CH
CH
CH
CH
CH
N
H
H
H
H
H
H
H
H
H
H
H
H
H
H
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
C
CH
CH
CH
CH
CH
CH
CH
CH
CH
CH
CH
CH
CH
CH
CH
CH
N
CH
CH
CH
CH
CH
CH
CH
CH
CH
CH
CH
CH
CH
CH
CH
CH
CH
CH
N
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
3-F
4-OMe
4-OMe
4-Cl
4-Br
4-Cl
4-Cl
4-Cl
4-Cl
4-Cl
4-Cl
4-Cl
4-Cl
4-Cl
4-Cl
4-Cl
4-Cl
4-Cl
4-Cl
4-Cl
4-Cl
35
20
7
6
5
Pyrrolidin-1-yl
4-Methyl-piperidin-1-yl
4-Aminocarbonyl-piperidin-1-yl
4-Trifluoromethyl-4-hydroxy-piperidin-1-yl
4-Methylpiperazin-1-yl
Pyrrolidin-1-yl
Pyrrolidin-1-yl
Pyrrolidin-1-yl
Pyrrolidin-1-yl
Pyrrolidin-1-yl
Pyrrolidin-1-yl
Pyrrolidin-1-yl
Pyrrolidin-1-yl
Pyrrolidin-1-yl
Pyrrolidin-1-yl
Pyrrolidin-1-yl
4-Methyl-piperidin-1-yl
8
6
10
11
5
26
6
10
77
20
5
78
12
17
4
CH
N
N
CH
CH
CH
CH
CH
CH
CH
CH
CH
CH
CH
CH
2-Cl
3-Cl
H
H
H
N
N
N
N
CH
CH
O
O
H
CH
a
For the description of biological methods, see Ref. 7.
selectivity was achievable. Polarity in the amine part appears to
influence affinity to the 5-HT2A receptor more than to the M1
receptor (46, 61), a substituent in the left hand aromatic system
reduced affinity to the M1 receptor (62 vs 61) whereas a sub-
stituent in the right hand aromatic system seem to increase affinity
to the M1 receptor (63 vs 45). But overall a clear structure–selec-
tivity relationship could not be established in this series. In con-
Further in vitro investigations of 43 revealed that it is not a P-gp
substrate and therefore a reasonable brain exposure (as indicated
by cerebrospinal fluid (CSF) levels) was detected in pharmacoki-
netic investigations (Fig. 3).
Based on these encouraging results we used 43 (=BI 414) to
investigate the influence of pharmacological MCH-R1 antagonism
on food intake and body weight in an animal model of obesity.
Our model consists of three animal groups (rat, exbreeder, control
group fed with standard rat chow diet, two groups with access to
chow and high fat diet for the whole experiment) and two different
compound treatment periods (43, 15 and 7.5 mg/kg per oral, once
trast to the amide series, the alkyne series showed
a less
pronounced interaction potential for CYP enzymes (e.g., CYP2D6)
and an overall acceptable metabolic stability, thus qualifying this
series for further in vivo investigations.

3268
S. G. Müller et al. / Bioorg. Med. Chem. Lett. 25 (2015) 3264–3269
Table 5
Optimization of alkyne series towards improved selectivity
Y
X
N
2
R
N
1
R
O
W
R1R2N
W
X
Y
MCH-R1a
IC50 (nM)
M1
IC50 (nM)
5HT2A
IC50 (nM)
CYP 2D6
HLM/RLM
t1/2 (min)
IC50
(lM)
43
45
46
61
62
63
Pyrrolidin-1-yl
4-Methyl-piperidin-1-yl
4-Aminocarbonyl-piperidin-1-yl
(R)-2-Hydroxymethyl-pyrrolidin-1-yl
(R)-2-Hydroxymethyl-pyrrolidin-1-yl
4-Methyl-piperidin-1-yl
H
H
H
H
Me
H
H
H
H
H
H
F
Cl
Cl
Cl
Cl
Cl
Cl
7
5
8
8
4
4
170
500
1354
>10,000
3290
1640
6
>90/>68
>90/22
>90/21
NT
>90/>45
NT/41
2210
>10000
285
2890
665
>10
>50
NT
>50
>50
>10,000
a
For the description of biological methods, see Ref. 7; NT: not tested; HLM: human liver microsomes; RLM: rat liver microsomes.
dose
Cplasma@2h
[nM]
CCSF@2h
[nM]
Ki
CCSF/Ki@2/8/24h
1.2/0.5/0.1
[µmol/kg]
[nM]
43
56.9
256
5.5
4
Figure 3. In vivo data 43 after oral application in Wistar rats.
Figure 4. In vivo efficacy of compound 43 in animal model of obesity.
daily). The results obtained for the different groups are summa-
rized in Figure 4. The weight gain in the animal groups on high
fat diet (feeding period) could be reversed with compound treat-
ment (period 1). The reduction of the dose resulted in a weight
stable period for the animals in the treated group (period 2). A
rebound in weight gain was seen after the treatment stopped.
Compound 43 was well tolerated over the whole treatment period
without any overt signs of adverse effects. To confirm these results
the more selective8 compound 45 was tested in the same setting
(treatment period 1) resulting in a comparable outcome with
respect to body weight reduction and tolerability (data not shown).
These results lend strong support to the concept of MCH-1R
antagonism for the treatment of obesity. Unfortunately, 43, 45
and other compounds from the alkyne series could not be pursued
further, as a rather strong hERG interaction and phospholipidosis
potential (both determined in vitro) were observed throughout

S. G. Müller et al. / Bioorg. Med. Chem. Lett. 25 (2015) 3264–3269
3269
serum albumin (protease-free), 0.021% bacitracin, 1
leupeptin and phosphoramidone) to a concentration of 5–15
200 L of this membrane fraction (contains 1–3 lg of protein) are incubated for
l
g/ml aprotinin, 1
l
g/ml
the series. The efforts to overcome these liabilities with new series
of MCH-1R antagonists are described in the subsequent Letters.9,10
1
lM
lg/ml.
l
60 min at ambient temperature with 100 pM of 125I-tyrosyl melanin
concentrating hormone (125I-MCH commercially obtainable from NEN) and
Acknowledgments
increasing concentrations of the test compound in a final volume of 250 lL.
After the incubation the reaction is filtered using a cell harvester through 0.5%
PEI treated fibreglass filters (GF/B, Unifilter Packard). The membrane-bound
radioactivity retained on the filter is then determined after the addition of
scintillator substance (Packard Microscint 20) in a measuring device (TopCount
of Packard). The non-specific binding is defined as bound radioactivity in the
We thank Nicole Mayer, Stefanie Rausenberger, Siegfried Kolb,
Markus Schranz and Klaus Heinrich for their excellence in prepar-
ing the compounds shown in this manuscript.
presence of 1 lM MCH during the incubation period. The analysis of the
References and notes
concentration binding curve is carried out on the assumption of one receptor
binding site. IC50 values are the mean of at least two separately performed
experiments. Standard: Non-labelled MCH competes with labelled 125I-MCH
for the receptor binding with an IC50 value of 0.06–0.15 nM. The KD value of the
radioligand is 0.16 nM.
1. Grundy, S. M. Nat. Rev. Drug Disc. 2006, 51, 295.
2. Chung, S.; Parks, G. S.; Lee, C.; Civelli, O. J. Mol. Neurosci. 2011, 43, 115.
3. Johannson, A. Expert Opin. Ther. Patents 2011, 21, 905.
8. >100 fold selective versus 43 GPCR’s, 5 ion channels and 2 transporter (CEREP
panel).
9. Roth, G. J.; Heckel, A.; Kley, J. T.; Lehmann, T.; Müller, S. G.; Oost, T.; Rudolf, K.;
Arndt, K.; Budzinski, R.; Lenter, M.; Lotz, R. R. H.; Schindler, M.; Thomas, L.;
Stenkamp, D. Bioorg. Med. Chem. Lett. 2015, 25, 3270.
10. Oost, T.; Heckel, A.; Kley, J. T.; Lehmann, T.; Müller, S. G.; Roth, G. J.; Rudolf, K.;
Arndt, K.; Budzinski, R.; Lenter, M.; Lotz, R. R. H.; Maier, G.-M.; Markert, M.;
Thomas, L.; Stenkamp, D. Bioorg. Med. Chem. Lett. 2015, 25, 3275.
4. Reasor, M. J.; Hastings, K. L.; Ulrich, R. G. Expert Opin. Drug Saf. 2006, 5, 567.
5. Méndez-Andino, J. L.; Wos, J. A. Drug Discovery Today 2007, 12, 972.
6. Högberg, T.; Frimurer, T. M.; Sasmal, P. K. Bioorg. Med. Chem. Lett. 2012, 22,
6039.
7. Müller, S. G.; Stenkamp, D.; Arndt, K.; Roth, G. J.; Lotz, R. H.; Lehmann-Lintz, T.;
Lenter, M.; Lustenberger, P.; Rudolf, K. WO 04039780, 2004. MCH-R1 binding
assay: Membranes from CHO/Galpha16 cells stably transfected with human
MCH-R1 are re-suspended using a syringe (needle 0.6 Â 25 mm) and diluted in
test buffer (50 mM HEPES, 10 mM MgCl2, 2 mM EGTA, pH 7.00; 0.1% bovine