Serendipitous discovery of a new class of agonists for the melanocortin 1 and 4 receptors and a new class of cyclophanes

Article abstract of DOI:10.1016/j.bmcl.2011.01.011

A new class of melanocortin 4 receptor (MC4r) agonists was discovered from an unexpected sidereaction in which formaldehyde caused cyclization. These cyclophanes were found to be sub micromolar agonists of the MC1 and MC4 and were less potent on the MC3 and MC5 receptor. They were shown to compete with the peptidic antagonist SHU9119 for binding to the MC4 receptor. In an acute feeding study in Sprague Dawley rats, food intake was reduced more than 50% versus vehicle after 3 h at a dose of 1 mg/kg.

Full text of DOI:10.1016/j.bmcl.2011.01.011

Bioorganic & Medicinal Chemistry Letters 21 (2011) 1459–1463
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
Serendipitous discovery of a new class of agonists for the melanocortin
1 and 4 receptors and a new class of cyclophanes
Kilian Conde-Frieboes a, , Michael Ankersen b, Jens Breinholt c, Birgit Sehested Hansen d, Kirsten Raun e,
⇑
Henning Thøgersen f, Birgitte S. Wulff g
a Department of Protein & Peptide Chemistry, Novo Nordisk, Novo Nordisk Park, 2760 Måløv, Denmark
b CMR—Clinical Pharmacology Research, Novo Nordisk Inc., 100 College Road West, Princeton, NJ 08540, USA
c Protein Structure and Biophysics, Novo Nordisk, Novo Nordisk Park, 2760 Måløv, Denmark
d Clinical Pharmacology, Novo Nordisk, 2860 Søborg, Denmark
e Department of Pharmacology, Novo Nordisk, Novo Nordisk Park, 2760 Måløv, Denmark
f Structure, Novo Nordisk, Novo Nordisk Park, 2760 Måløv, Denmark
g Department of GLP-1 & Obesity Biology, Novo Nordisk, Novo Nordisk Park, 2760 Måløv, Denmark
a r t i c l e i n f o
a b s t r a c t
Article history:
A new class of melanocortin 4 receptor (MC4r) agonists was discovered from an unexpected sidereaction
in which formaldehyde caused cyclization. These cyclophanes were found to be sub micromolar agonists
of the MC1 and MC4 and were less potent on the MC3 and MC5 receptor. They were shown to compete
with the peptidic antagonist SHU9119 for binding to the MC4 receptor. In an acute feeding study in Spra-
gue Dawley rats, food intake was reduced more than 50% versus vehicle after 3 h at a dose of 1 mg/kg.
Ó 2011 Elsevier Ltd. All rights reserved.
Received 16 November 2010
Revised 3 January 2011
Accepted 4 January 2011
Available online 7 January 2011
Keywords:
Melanocortin receptor 4
Agonist
Cyclophane
Combinatorial chemistry
Obesity
The melanocortin receptors constitute a group of five receptors
(MC1R-MC5R) belonging to the family A of seven transmembrane
G-protein coupled receptors.1,2 The endogenous ligands are de-
rived from proopiomelanocortin (POMC) and are small peptides.3,4
In addition, two endogenously expressed antagonists, agouti and
agouti gene related peptide (AgRP), are belonging to the system.5
The MC3 and MC4 receptors have been linked to sexual behavior
and energy homeostasis.6–8 Mutations in the MC4 receptor are
coupled to obesity in humans,9 and in rodents subchronic treat-
ment with the synthetic peptide ligand MT-II (1) resulted in weight
research program focused on the discovery and development of or-
ally available MC4R agonists was initiated.
In essence the overall approach was to synthesize small mole-
cule libraries based on in silico design and subsequently screen
the libraries for MC4R agonistic activity. Thus, using the modeling
17
package SYBYL a model of MT-II was overlayed with new scaffold
candidates containing the pharmacophores. Scaffolds which were
predicted to spatially present the pharmacophores in a manner
that allowed for strong receptor interactions were identified. Small
focused libraries were synthesized using solid phase technology
and after cleavage from the solid support the crude compounds
were without further purification screened for MC4R agonistic
activity using a cell based assay. MC4R was stably transfected into
baby hamster kidney (BHK) cells with a reporter system consisting
of a cyclic-AMP responsive element (CRE) coupled to a luciferase
reporter gene.18 Responses were compared to the effect of MT-II
and expressed relative (in %) to the maximum activity of MT-II
(Emax). MT-II was considered to be a full agonist at the MC4R. Com-
pounds identified by this method were further characterized in
binding and cAMP assays for other relevant MC receptors.
loss.10
a-MSH (2), which is one of the endogenous ligands for the
MC1, MC3, MC4 and MC5 receptors, is a tridecapeptide. Struc-
ture–activity analysis by Hruby et al. revealed the pharmacophores
important for the interaction between a-MSH and the melanocor-
tin receptors.11–14 Scheme 1 depicts MT-II and
pharmacophores emphasized.
a-MSH with the
Despite the urgent and growing need for a pharmacological
treatment of obesity very few effective treatment options are cur-
rently available.15,16 With the aim to satisfy this unmet need, a
MC4R agonistic activity was detected in a library based on a scaf-
fold carrying four pharmacophores and synthesized using simple
acylationand reductivealkylationchemistry asoutlinedinScheme2.
⇑
Corresponding author. Tel.: +45 44434880.
E-mail address: kcf@novonordisk.com (K. Conde-Frieboes).
0960-894X/$ - see front matter Ó 2011 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2011.01.011

1460
K. Conde-Frieboes et al. / Bioorg. Med. Chem. Lett. 21 (2011) 1459–1463
O
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
O
O
O
O
O
O
O
O
O
1
N
N
O
N
O
O
O
O
O
O
O
N
O
O
O
N
N
N
O
N
N
N
N
N
N
N
N
N
N
O
O
O
O
O
N
O
N
S
N
2
N
Scheme 1. MT-II (1) and
a-MSH (2). The important pharmacophores are shown in red.
O
R1
O
Wng Pol
R1
Wng Pol
O
N
N
O
N
N
R2
O
O
O
O
R1
R2
R4
R1
WngPol
N
Wng Pol
N
O
N
N
O
N
N
PG
R3
R3
O
R2
O
R1
R4
N
N
N
R3
O
R2
Scheme 2. Outline of the synthesis.
The library consisted of 384 random compounds out of 7776
possible (6 Â 9 Â 8 Â 18), which covers 5% of the chemical space
represented by these building blocks and was synthesized in our
Parallel Synthesis Lab using a ACT384 robot. According to the li-
brary design, the active compound was expected to have the struc-
ture 3 shown in Scheme 3.
The crude material exhibiting MC4R agonistic activity was sub-
jected to LC–MS analysis. Surprisingly, the LC–MS data demon-
strated that the major component had a molecular mass of
655 Da corresponding to a net mass increase of 42 Da compared
to the mass of the expected compound (3) (613 Da). Also present
in the mixture was a minor component exhibiting a surplus mass
of 54 Da compared to that of structure 3. Thus the screening hit
was of unknown structure formed by an unknown side reaction.
By manual synthesis using the same chemistry as applied in the
automated library synthesis compound 3 was indeed the major
product formed as evidenced by LC–MS analysis (613 Da). Com-
pound 3 was devoid of any activity in the MC4R agonist assay.
The active crude material from the library synthesis was sub-
jected to HPLC-fractionation and a fraction accounting for the
MC4R agonistic activity was collected. The NMR spectrum of the
active fraction indicated the presence of several components and
could not be analyzed in depth. However, the NMR data served
to rule out simple interpretations such as isopropylation or acety-
lation for the mass difference of 42 Da.
An important hint to the nature of the +42 Da components and
the underlying chemistry was the fact that all library samples
carrying a tyrosine in the R3 position contained components with
increased molecular masses of 42 Da. One possible explanation
for the presence of the +42 Da products is a reaction between
formaldehyde of an yet unknown source and a tyrosine residue
to form 4H-benzo[1,3]dioxin compounds. Thus by reaction with
O
O
O
O
N
N
N
N
N
N
NH2
N
NH2
N
HO
O
O
O
O
3
4
Scheme 3. Intended structure 3 and 4H-benzo[1,3]dioxin 4.

K. Conde-Frieboes et al. / Bioorg. Med. Chem. Lett. 21 (2011) 1459–1463
1461
formaldehyde compound 3 could form compound 4 (Scheme 3)
with a molecular mass of 655 Da.
5 demonstrated that the fragment originating from cleavage of the
amide bond between the R3–R4 moiety and the R1–R2 fragment, are
observed in the spectra of 5 but are absent in those recorded for the
+54 Da component. This is in accordance with the presence of an
additional stable bond tethering the two fragments as illustrated
in 6. NMR data were in accordance with the proposed overall struc-
ture 6 and demonstrated the presence of both the para- and the
ortho-isomer in a 2:1 ratio. Finally, the structure (6) was unambig-
uously proven by synthesis of the para-isomer (6a) using a ring clo-
sure reaction under acidic conditions (Scheme 6).
To challenge this hypothesis, compound 3 was synthesized
using the manual procedure described with the only exception that
paraformaldehyde was deliberately added during the TFA cleavage
step. This experiment provided a product exhibiting MC4R agonist
activity and the major product was indeed the desired 4H-ben-
zo[1,3]dioxin 4 corresponding to the +42 Da product of compound
3. Interestingly, in this sample produced by the deliberate addition
of formaldehyde to the cleavage step, the +42 Da compound 4 was
accompanied by a minor +54 Da component which was also ob-
served in the active library sample. This indicated that the
+54 Da component was also formaldehyde derived.
Using the chemistry including formaldehyde, a small library (24
compounds) of 4H-benzo[1,3]dioxin compounds was synthesized
and screened for MC4R agonistic activity without purification after
cleavage from the solid support. The data implied that compounds
where R4 was a methoxyacetyl group were more potent than the
cyclopropyl analogues. Thus, the next step was to synthesize com-
pound 5 (the methoxyacetyl analogue of 4) under more controlled
conditions. The synthesis of 5 is outlined in Scheme 4.
Surprisingly, the purified compound 5 also tested negative in
the MC4R agonist assay. Now the additional +54 Da minor byprod-
ucts observed both in the original active library sample and in the
product obtained when including formaldehyde in the manual
synthesis of 3 came under scrutiny. Using a chiral analytical HPLC
column (Chiralpak AD 250 Â 4.6 mm, heptane/ethanol + 0.1% DEA
1:4) it was possible to separate the +54 Da component from the
main compound 5. The +54 Da component was found to account
for the MC4R agonistic activity and structure 6 was assigned to this
compound (Scheme 5).
Structure 6a was simplified to S-56-Methoxy-9-(3-(4-(3-amino-
propyl) piperazin-1-yl)propyl)-7-(2-methoxyacetylamino)-1(1,3),
3(1,4),5(1,3)-tribenzena-2-oxa-9-aza-cyclodecaphane (7),19,20 syn-
thesized using similar conditions. The NMR spectra of 6a and 7
both exhibited duplication of signals due to the presence of amide
bond rotamers and line broadening attributable to ring flipping of
the piperazine ring. Additionally, individual resonances were
observed for each of the four protons in the p-substituted ring
indicative of hindered ring rotation, a phenomenon also observed
in similar systems, for example, acerogenin.21 While the signal
duplication due to hindered rotation around the amide bond
disappeared upon heating to 120 °C, degeneration of the signals
O
O
O
N
N
N
N
N
O
O
O
Both MS and NMR analysis provided compelling evidence sup-
porting the proposed structure 6 for the +54 Da component. Com-
parison of LC–MS/MS data for the +54 Da component with those of
6
Scheme 5. Proposed structure 6.
O
O
H
O
O
H
i
iii
ii
H2N
H
N
N
O
N
O
O
O
OH
O
O
O
O
O
HO
O
O
HO
O
O
O
iv
H2N
N
H
H
HN
O
N
+
N
N
H
Boc
N
N
Boc
v
vi
O
O
H
N
N
N
O
N
NH2
O
O
O
O
5
Scheme 4. Synthesis of compound 5. Reagents and conditions: (i) 1 equiv HOBt, 1 equiv TBTU, 1.5 equiv DIPEA; (ii) TFA/DCM 1:1, 2 equiv paraformaldehyde, 1 h; (iii) LiOH,
H2O. (iv) ethanol, Raney-Ni, H2; (v) 1.2 equiv 2,4,6-tripropyl-1,3,5,2,4,6-trioxatriphosphorinane-2,4,6-trioxide (PPAA), 5.9 equiv N-methyl morpholin, DCM; (vi) TFA.

1462
K. Conde-Frieboes et al. / Bioorg. Med. Chem. Lett. 21 (2011) 1459–1463
O
O
N
O
N
N
N
O
O
N
O
N
O
O
i
N
N
+
Troc
O
N
N
O
Troc
O
O
O
O
O
O
O
O
O
O
O
O
O
N
N
N
N
ii
iii
N
N
N
N
N
N
O
O
O
O
O
6a
7
Scheme 6. Synthesis of macrocycle 6a and structure of macrocycle 7. Reagents and conditions: (i) 1.5 equiv collidine, 1.5 equiv HATU, DCM; (ii) TFA/DCM 1:1 90 min; (iii) Zn,
HOAc.
Table 1
150
100
50
Binding data for compounds 6a and 7 (mean SEM, n)
hMC1 Ki (nM)
hMC3 Ki (nM)
hMC4 Ki (nM)
hMC5 Ki (nM)
1000 nM SHU9119
100 nM SHU9119
10 nM SHU9119
0 nM SHU9119
6a
7
328 16, 3
116 12, 3
4530 1846, 3
1182 360, 3
403 171, 3
392 69, 3
1532 364, 3
388 46, 3
corresponding to protons in the p-substituted aromatic ring was
not observed.
The macrocycle 7 was further tested in vitro in binding and
cAMP assays and for effect on food intake in vivo. Tables 1 and 2
show binding and functional data for compounds 6a and 7 for
the relevant melanocortin receptors. Both compounds are agonists
for the MC1 and MC4 receptors and do not show any detectable
activation of the MC3 and MC5 receptors (Table 1 and 2). Using
the established peptidic MC4 receptor antagonist SHU911922 it is
possible to show competitive antagonism, indicating a similar
binding site for the peptide SHU9119 and the macrocycle 6a (
Fig. 1).
0
-14 -13 -12 -11 -10
-9
-8
-7
-6
-5
-4
-3
Concentration of 6a (-log(M))
Figure 1. Inhibition of the agonist 6a with 10, 100 or 1000 nM of the antagonist
SHU9119. The maximal stimulation with compound 6a is set to 100% and the other
result is calculated as relative values to maximal stimulation with compound 6a.
Compound 7 was tested in an acute feeding model using male
Sprague Dawley rats and was able to reduce food intake with a po-
tency comparable to or better than Sibutramine (Fig. 2) in the first
3 h after ip injection.
The source of the formaldehyde is most likely the rack in which
the samples were stored after cleavage and evaporation of the tri-
fluoroacetic acid. The racks were composed of Bakelite, a phenol/
formaldehyde polymer. In conclusion, a novel class of cyclophanes
was discovered as potent and selective MC4R agonists. The active
components were formed by an unintended side reaction involving
formaldehyde. Despite an in vitro activity of around a half a
Attempts to obtain a structure–activity relationship (SAR) were
not successful. Changes in the aromatic ring on the amino acid (R3)
lead to structural isomers which were usually hard to separate,
while replacement of the amine (R1) with shorter and simpler
amines lead to inactive compounds, thus indicating the importance
of a having certain distance between the cyclophane and the termi-
nal amine.
Table 2
Functional data for compounds 6a and 7 (mean SEM, n)
hMC1 EC50 (nM)
hMC1 Emax (%)
hMC3 EC50 (nM)
hMC4 EC50 (nM)
hMC4 Emax (%)
hMC5 EC50 (nM)
6a
7
12 1, 3
16 3, 3
85
83
4
1
>10,000, 3
>10,000, 3
697 138, 13
466 56, 16
98 5, 13
61 5, 16
>10,000, 3
>10,000, 3

K. Conde-Frieboes et al. / Bioorg. Med. Chem. Lett. 21 (2011) 1459–1463
1463
15
10
5
include MOL files and InChiKeys of the most important compounds
described in this article.
References and notes
**
**
*
**
1. Cone, R. D. The Melanocortin Receptors, 1st ed.; Humana Press: Totowa, New
Jersey, 2000.
2. Lerner, A. B. Ann. N. Y. Acad. Sci. 1993, 680, 1.
**
*
**
**
**
2h 3h
1h 2h 3h
3h
3h
1h
1h
3h
1h 2h
1h 2h
3. Harris, J. I.; Lerner, A. B. Nature 1957, 179, 221.
2h
0
4. Adan, R. A. H.; Gispen, W. H. Eur. J. Pharmacol. 2000, 405, 13.
5. Ollmann, M. M.; Wilson, B. D.; Yang, Y. K.; Kerns, J. A.; Chen, Y.; Gantz, I.; Barsh,
G. S. Science 1997, 278, 135.
Vehicle
Sibutramine
3 mg/kg
7
7
7
0.1 mg/kg 0.3 mg/kg
1 mg/kg
6. Butler, A. A.; Marks, D. L.; Fan, W.; Kuhn, C. M.; Bartolome, M.; Cone, R. D. Nat.
Neurosci. 2001, 4, 605.
7. Butler, A. A.; Kesterson, R. A.; Khong, K.; Cullen, M. J.; Pelleymounter, M. A.;
Dekoning, J.; Baetscher, M.; Cone, R. D. Endocrinology 2000, 141, 3518.
8. Pfaus, J. G.; Shadiack, A.; Van Soest, T.; Tse, M.; Molinoff, P. Proc. Natl. Acad. Sci.
U.S.A. 2004, 101, 10201.
Figure 2. In vivo effect on food intake in scheduled fed male Sprague Dawley rats,
which have been trained to consume a 24 h food ration during 5 h. Macrocycle 7
(0.1, 0.3 and 1 mg/kg) was administered ip to 8 rats just before food was presented.
Sibutramine16 (3 mg/kg) was used as positive control. Food intake was followed for
3 h and compared to a vehicle treated group (⁄p <0.05, ⁄⁄p <0.01, ⁄⁄⁄p <0.001). A
mixture of 20% Tween 20, 20% propylene glycol and 50 mM phosphate buffer pH 6.0
was used as vehicle.
9. Farooqi, I. S.; Keogh, J. M.; Yeo, G. S.; Lank, E. J.; Cheetham, T.; O’Rahilly, S. N.
Engl. J. Med. 2003, 348, 1085.
10. Hamilton, B. S.; Doods, H. N. Obes. Res. 2002, 10, 182.
11. Hruby, V. J.; Lu, D.; Sharma, S. D.; Castrucci, A. L.; Kesterson, R. A.; al Obeidi, F.
A.; Hadley, M. E.; Cone, R. D. J. Med. Chem. 1995, 38, 3454.
12. Haskell-Luevano, C.; Sawyer, T. K.; Hendrata, S.; North, C.; Panahinia, L.; Stum,
M.; Staples, D. J.; Castrucci, A. M. D.; Hadley, M. E.; Hruby, V. J. Peptides 1996,
17, 995.
13. Haskell-Luevano, C.; Hendrata, S.; North, C.; Sawyer, T. K.; Hadley, M. E.; Hruby,
V. J.; Dickinson, C.; Gantz, I. J. Med. Chem. 1997, 40, 2133.
14. Haskell-Luevano, C.; Nikiforovich, G.; Sharma, S. D.; Yang, Y. K.; Dickinson, C.;
Hruby, V. J.; Gantz, I. J. Med. Chem. 1997, 40, 1738.
micromolar, the compounds displayed good in vivo potency. We
suspect this may be due to a high penetration rate of the blood
brain barrier. In conclusion, a novel class of cyclophanes was dis-
covered as potent and selective MC4R agonists.
Acknowledgments
15. Hauptman, J. B.; Jeunet, F. S.; Hartmann, D. Am. J. Clin. Nutr. 1992, 55, 309S.
16. Heal, D. J.; Aspley, S.; Prow, M. R.; Jackson, H. C.; Martin, K. F.; Cheetham, S. C.
Int. J. Obes. 1998, 22, S18.
17. SYBYL, Tripos International, 1699 South Hanley Rd., St. Louis, Missouri 63144,
USA.
The authors would like to thank Patrick W. Garibay for editorial
comments and Peter Andersen and Rie K. Schjeltved for assistance
with the synthesis work.
18. Fitzgerald, L. R.; Mannan, I. J.; Dytko, G. M.; Wu, H. L.; Nambi, P. Anal. Biochem.
1999, 275, 54.
19. Powell, W. H. Pure Appl. Chem. 1998, 70, 1513.
20. Favre, H. A.; Hellwinkel, D.; Powell, W. H.; Smith, H. A.; Tsay, S. S. C. Pure Appl.
Chem. 2002, 74, 809.
A. Supplementary data
21. Gonzalez, G. I.; Zhu, J. J. Org. Chem. 1999, 64, 914.
22. Schiöth, H. B.; Muceniece, R.; Mutulis, F.; Prusis, P.; Lindeberg, G.; Sharma, S. D.;
Hruby, V. J.; Wikberg, J. E. S. Peptides 1997, 18, 1009.
Supplementary data associated with this article can be found, in
the online version, at doi:10.1016/j.bmcl.2011.01.011. These data