The identification of GPR3 inverse agonist AF64394; The first small molecule inhibitor of GPR3 receptor function

Article abstract of DOI:10.1016/j.bmcl.2014.09.077

The identification of the novel and selective GPR3 inverse agonist AF64394, the first small molecule inhibitor of GPR3 receptor function, is described. Structure activity relationships and syntheses based around AF64394 are reported.

Full text of DOI:10.1016/j.bmcl.2014.09.077

Bioorganic & Medicinal Chemistry Letters 24 (2014) 5195–5198
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
The identification of GPR3 inverse agonist AF64394; The first small
molecule inhibitor of GPR3 receptor function
Thomas Jensen, Lisbeth Elster, Søren Møller Nielsen, Suresh Babu Poda, Frosty Loechel,
⇑
Christiane Volbracht, Ib Vestergaard Klewe, Laurent David, Stephen P. Watson
Neuroscience Drug Discovery, H-Lundbeck A/S, 2500 Valby, Copenhagen, Denmark
a r t i c l e i n f o
a b s t r a c t
Article history:
The identification of the novel and selective GPR3 inverse agonist AF64394, the first small molecule
inhibitor of GPR3 receptor function, is described. Structure activity relationships and syntheses based
around AF64394 are reported.
Received 11 September 2014
Revised 24 September 2014
Accepted 25 September 2014
Available online 2 October 2014
Ó 2014 Elsevier Ltd. All rights reserved.
Keywords:
GPR3
GPR6
GPR12
G protein coupled receptor
Orphan receptor
Inverse agonist
AF64394
GPR3 is orphan G-protein coupled receptor (GPCR) belonging to
a branch of the GPCR phylogeny that includes the Melanocortin and
Cannabinoid peptide receptors as well as the sphingophosphate
lipid receptors.1 The sequence similarity on the amino acid level
is however only close to that of the GPR6 and GPR12, two other
orphan receptors, which together with GPR3 thus makes a distinct
GPCR family.1a When clustered based on sequence similarity of the
transmembrane (TM) bundle binding site only, the GPR3/6/12
group clusters more broadly with the S1P/LPA (EDG) and prosta-
noid receptors.1b The GPR3/6/12 group of receptors is widely
believed to be constitutively active, resulting in elevated levels of
cellular cAMP when overexpressing these receptors in heterologous
expression systems as well as in primary neurons.2–4
that GPR3 did not contain a drugable ligand binding domain. Ear-
lier this year it was however demonstrated2 that Diphenyleneiod-
onium chloride (DPI) is a specific agonist of GPR3 with no apparent
activity on either GPR6 or GPR12 as well as on a range of well
described GPCRs including the alpha and beta-adrenergic
receptors.
Herein we report the identification of a series of GPR3 inverse
agonists, exemplified by AF64394. We anticipate these compounds
will serve as useful tools in enabling understanding of the role of
GPR3 in neurodegenerative and neuropsychiatric disease states,
and in further exploration of GPR3 as a drug target.
Cl
Intriguingly GPR3 activity has been found to be positively cou-
pled to Ab production from the processing of the amyloid pre-cursor
protein (APP)—perhaps through an interaction with the GPCR scaf-
folding protein bArrestin2 that has also been shown to be a factor in
the regulation of Ab production.5,6 Using mice genetically ablated
for GPR3, the receptor has been also been observed to be involved
in both pain sensation, anxiety and addictive behaviours.7–9
On a molecular level it has not been ruled out that it is an intrin-
sic factor from the cellular environment or the growth media that
O
HN
N
HN
N
N
N
N
N
N
N
I+
Cl
AF64394
1
DPI
GPR3 pIC50 = 7.3
GPR3 pIC50 = 5.4
drives the constitutive coupling to the Gas and thus the cellular
increase in cAMP. Also the possibility has until recently remained
From a GPR3 High Throughput Screen10 of the Lundbeck screening
collection [1,2,4]triazolo[1,5-a]pyrimidine (1) emerged as
a
promising hit. The compound demonstrated inverse agonism of
the constitutively activated GPR3 receptor, effecting a ca. 80%
⇑
Corresponding author.
http://dx.doi.org/10.1016/j.bmcl.2014.09.077
0960-894X/Ó 2014 Elsevier Ltd. All rights reserved.

5196
T. Jensen et al. / Bioorg. Med. Chem. Lett. 24 (2014) 5195–5198
R1
R2
N
N
N
Cl
NH2
O
N
N
N
Ar
O
i.
N
iii.
N
NH
N
+
ii.
Ar
N
Ar
N
OEt
3
2
4
Scheme 1. Reagents and conditions: (i) AcOH, reflux, (ii) POC13, reflux, (iii) R1R2NH, EtOH, rt or DCM, Et3N, rt or DMF, K2CO3, rt ? 35 °C.
R1
R2
N
Cl
N
N
N
NH2
NH
O
N
N
N
N
N
iii. R1R2NH2
iv. ArM
i.
EtO
O
N
+
ii.
Cl
Ar
N
OEt
5
2
Scheme 2. Reagents and conditions: (i) NaOEt, EtOH, reflux, (ii) POC13, N,N-dimethylaniline, reflux, (iii) R1R2NH, EtOH, rt (iv) Pd(dtbpf)Cl2, K3PO4, 1,4-dioxane/H2O,
120 °C, w.
l
R2
N
OH
Cl
HN
N
HCO2H
ii.
iii.
i.
N
N
N
N
N
N
Ph
N
NHNH2
Ph
N
Ph
8
6
7
Scheme 3. Reagents and conditions: (i) 60 °C (ii) POC13, 60 °C, (iii) R1NH2, EtOH, rt.
inhibition of cAMP synthesis with an associated IC50 of 4 lM
Table 1
GPR3 inverse agonism potency of compounds 9
(pIC50 = 5.4). The effect was confirmed to be GPR3 specific by
demonstrating no inhibition of cAMP levels stimulated by D1
receptor agonism in the same cell background with the same assay
technology.10
X
A programme of analogue synthesis based on compound 1 was
initiated with the aim of developing structure activity relationships
(SAR) and enhancing properties to afford a molecular probe for
GPR3.
The [1,2,4]triazolo[1,5-a]pyrimidines (2, 9, 13 and 14) were syn-
thesised according to the chemistry or Scheme 1 or Scheme 2 or
variations thereof.
b-Keto-esters (3) are condensed with 3-amino-1,2,4,triazole to
afford, after activation with POCl3, the 7-chloro-[1,2,4]triazolo[1,5-
a]pyrimidines (4).11 Simple chloride displacement then affords the
final compounds (2) (Scheme 1). Alternatively reaction of diethyl-
malonate with 3-amino-1,2,4,triazole affords, after activation with
POCl3, dichloride (5).12 Simple displacement of the 7-chloride of 5
followed by a palladium catalysed Suzuki cross-coupling affords
the final compounds (2) (Scheme 2).
The regioisomeric [1,2,4]triazolo[4,3-a]pyrimidines (8) were
prepared according to Scheme 3. The known hydrazinopyrimidone
(6) is condensed with formic acid to afford, after chlorination with
POCl3, the 5-Chloro-[1,2,4]triazolo[4,3-a]pyrimidine (7).13,14 Sim-
ple chloride displacement then affords the final compounds (8)
(Scheme 3). The triazole regiochemistry of compounds of type 8
HN
N
N
N
N
9
Compound
X
GPR310
pIC50
Compound
X
GPR310
pIC50
1
H
5.4
5.0
4.5
4.4
6.1
5.6
6.1
9g
9h
9i
9j
9k
9l
2-OMe
2-OEt
6.0
6.6
5.2
5.2
6.7
5.8
5.3
9a
9b
9c
9d
9e
9f
2-CN
3-CN
4-CN
2-Cl
3-Cl
4-Cl
3-OMe
4-OMe
2-NMe2
3-NMe2
4-NMe2
9m
potency enhancing benzyl 2 substituents (Fig. 1). Interestingly the
regioisomeric [1,2,4]triazolo[4,3-a]pyrimidines (8a and 8b) are
essentially inactive. Thus a simple nitrogen positional switch from
the corresponding potent[1,2,4]triazolo[1,5-a]pyrimidines 9d and
9h is potency destroying, SAR which are indicative of a highly opti-
mised stereoelectronic interaction with the receptor in this region
of the molecule. Similarly the corresponding pyrazolo16 analogue
(10) is also inactive. Introduction of a methyl group at the 2
position (11) leaves potency essentially unchanged, whereas
introduction of methyl group at the 6-position (12) causes a ca. 10
fold reduction in potency relative to the unsubstituted parent 9d.
Attempts to enhance potency via re-engineering of the
benzylamine region of the compound 1 template via nitrogen
substitution, cyclisation, conformational restraint, or replacement,
and type
2 were confirmed and distinguished by NMR
techniques.15
Table 1 shows the effect of simple substitution of the benzyl
ring of compound 1 on the potency of GPR3 inverse agonism. Intro-
duction of an electron withdrawing cyano substituent is detrimen-
tal to potency in all positions. Other substituents show a modest
effect on potency in the benzyl 3 and 4 positions, but 2 substitu-
tions confers a clear potency enhancement, for example, in com-
pounds 9d and 9k.
The effect of changes to the central triazolo-pyrimidine fused-
ring on GPR3 potency was explored in examples bearing otherwise

T. Jensen et al. / Bioorg. Med. Chem. Lett. 24 (2014) 5195–5198
5197
X
O
Cl
HN
N
HN
N
HN
N
X
N
N
N
N
N
N
N
Y
N
11 X = H Y = Me pIC50 5.8
12 X = Me Y = H pIC50 5.2
8a X = Cl
8b X = OEt GPR3 pIC50 <4.5
GPR3 pIC50 <4.5
10 GPR3 pIC50 <4.5
Figure 1. GPR3 inverse agonism potency of compounds with triazolo-pyrimidine modifications.
Table 2
Table 3
GPR3 inverse agonism potency of compounds 2
GPR3 inverse agonism potency of compounds 13
X
N
N
N
Cl
N
HN
2
N
N
Compound
X
GPR310 Compound
pIC50
X
GPR310
pIC50
N
R
N
13
Compound
R
GPR310
pIC50
Compound
R
GPR310
pIC50
1
5.4
5.3
5.3
2f
5.8
4.5
4.8
H
N
*
N
*
9d
Ph
6.1
5.2
5.5
5.1
5.4
6.1
5.4
13g
13h
13i
13j
13k
13l
2-OMe
3-OMe
4-OMe
2-Pyridyl
3-Pyridyl
4-Pyridyl
Me
5.7
5.9
5.4
5.1
4.8
4.9
5.0
5.0
13a
13b
13c
13d
13e
13f
2-CN
3-CN
4-CN
2-Me
3-Me
4-Me
Cl
2a
2b
2g
2h
13m
13n
Me
N
*
N
*
Cyclopropyl
potency (Table 3). The majority of 5-aryl modifications caused a
fall in potency relative to comparator compound 9d, but with a
pattern of 3-aryl substitution being most benign. All 3 of the pyri-
dine substitutions (13j, k and l) caused a ca. 10-fold fall in potency,
and the corresponding low potency of the corresponding methyl
and cyclopropyl compounds (13m and n) demonstrated the impor-
tance of a 5-aryl substituent.
Cl
H
N
N
N
Me
*
*
*
H
N
Based on the SAR described in Tables 1–3 a further set of com-
pounds (14) was prepared with the objective of increasing GPR3
potency (Table 4) Introduction of a second chlorine atom at the
benzyl 4 position of compound 9d afforded a marginal increase
in potency (14a). However in combination with a 2-i-propoxy
group the 4-chloro substituent afforded a significant increase in
potency in the form of AF64394.17 Interestingly the 4-chloro sub-
stituent had the opposite effect when incorporated into morpho-
line compound 14b, affording the marginally less potent
compound 14c.
The selectivity of the most potent compounds of type 9 and 14
was determined against the closely related (see Introduction)
receptors GPR6 and GPR12. All compounds show at least 30 fold
selectivity over these receptors (see Table 5).
Compound 9h was also screened in a multi-target panel at
CEREP.18 The compound demonstrated some modest activity against
a small number of proteins/targets but was broadly selective.
In summary, in inverse agonist AF64394 and related com-
pounds we have identified the first true small molecule ligands
for GPR3 and the first inhibitors of GPR3 receptor function. The
compounds are specific and selective and thus form useful tools
for delineating the biology of GPR3. Whilst AF34394 and related
*
2c
<4.5
2i
4.6
H
Cl
Cl
2d
5.2
4.6
2j
4.7
4.9
H
H
N
*
O
*
2e
2k
N
*
*
Racemate.
were largely unsuccessful (Table 2 compounds 2a–i). Indeed only
in the tetrahydroisoquinoline (2f) was potency slightly increased.
Similarly modifications to the triazolo-pyrimidine 5-aryl group
were explored without achieving any significant enhancement in

5198
T. Jensen et al. / Bioorg. Med. Chem. Lett. 24 (2014) 5195–5198
Table 4
References and notes
GPR3 inverse agonism potency of further compounds 9
1. (a) Joost, P.; Methner, A. Genome Biol. 2002, 3, 11; (b) Gloriam, D. E.; Foord, S.
Ar
M.; Blaney, F. E.; Garland, S. L. J. Med. Chem. 2009, 52, 4429.
2. Ye, C.; Zhang, Z.; Wang, Z.; Hua, Q.; Zhang, R.; Xie, X. J. Pharmacol. Exp. Ther.
2014, 349, 437.
3. Tanaka, S.; Ishii, K.; Kasai, K.; Yoon, S. O.; Saeki, Y. J. Biol. Chem. 2007, 282,
10506.
HN
N
N
4. Uhlenbrock, K.; Gassenhuber, H.; Kostenis, E. Cell. Signal. 2002, 14, 941.
5. Nelson, C. D.; Sheng, M. PloS One 2013, 8, e74680.
N
N
6. Thathiah, A.; Horré, K.; Snellinx, A.; Vandewyer, E.; Huang, Y.; Ciesielska, M.; De
Kloe, G.; Munck, S.; De Strooper, B. Nat. Med. 2013, 19, 43.
7. Tourino, C.; Valjent, E.; Ruiz-Medina, J.; Herve, D.; Ledent, C.; Valverde, O. Br. J.
Pharmacol. 2012, 167, 892.
14
Compound Ar
GPR310 Compound Ar
pIC50
GPR310
pIC50
8. Ruiz-Medina, J.; Ledent, C.; Valverde, O. Neuropharmacology 2011, 61, 43.
9. Valverde, O.; Célérier, E.; Baranyi, M.; Vanderhaeghen, P.; Maldonado, R.;
Sperlagh, B.; Vassart, G.; Ledent, C. PloS One 2009, 4, e4704.
10. Compound potencies were determined by measuring changes in intracellular
cAMP (GPR3-R coupled to Gs). Time resolved fluorescence from Cisbio (cAMP
Cl
HTRFÒ
)
was applied to measure cAMP. This technique is based on
competitive immunoassay using cryptate-labeled anti-cAMP antibody and
d2-labeled cAMP. A 384-well assay format with a total assay volume of 20
a
Cl
9d
6.1
6.4
6.3
6.1
14d
7.0
7.3
6.8
7.0
*
O
O
O
O
lL
was used. 2500 HEK293-GPR3 (transfected cells GPR3 inducible cell line) cells
were incubated with sample compounds for 30 min at 37 °C using DPBS
containing 0.5 mM IBMX as stimulation buffer. After the addition of HTRFÒ
detection reagents signals at 620 and 665 nm (raw counts: ratio of 665:620)
were detected after a 3 h incubation. Values were calculated by nonlinear
regression using the sigmoid concentration–response (variable slope) using
Xlfit 4 (IDBS, UK). All values reported are average of at least 4 determinations.
GPR6 and 12 potencies were similarly determined.
*
Cl
Cl
14a
14b
AF64394
Cl
N
*
Cl
*
11. EP 0728759, 1996, p 40.
12. Gujjar, R.; El Mazouni, F.; White, K. L.; White, J.; Creason, S.; Shackleford, D. M.;
Deng, X.; Charman, W. N.; Bathurst, I.; Burrows, J.; Floyd, D. M.; Matthews, D.;
Buckner, F. S.; Charman, S. A.; Phillips, M. A.; Rathod, P. K. R. J. Med. Chem. 2011,
54, 3935.
14e
O
O
*
13. Attaby, F. A.; Eldin, S. M. Z. Naturforsch., B 1999, 54, 788.
14. Reimlinger, H.; Jacquier, R.; Daunis, J. Chem. Ber. 1971, 104, 2702.
15. Compound 9h: 1H NMR (600 MHz, DMSO-d6) d 8.75 (br s, 1H), 8.51 (s, 1H),
8.14–8.05 (m, 2H), 7.57–7.48 (m, 3H), 7.31 (dd, J = 7.5, 1.4 Hz, 1H), 7.28–7.20
(m, 1H), 7.03 (d, J = 8.2 Hz, 1H), 6.92–6.87 (m, 2H), 4.74 (d, J = 3.9 Hz, 2H), 4.12
(q, J = 7.0 Hz, 2H), 1.36 (t, J =7.0 Hz, 3H). Compound 8b. 1H NMR (600 MHz,
DMSO-d6) d 9.33 (s, 1H), 8.74 (br s, 1H), 8.32–8.00 (m, 2H), 7.64–7.45 (m, 3H),
7.38 (dd, J = 7.5, 1.6 Hz, 1H), 7.29 (td, J = 8.2, 1.7 Hz, 1H), 7.06 (d, J = 8.2 Hz, 1H),
6.94 (td, J = 7.5, 0.9 Hz, 1H), 6.66 (s, 1H), 4.72 (s, 2H), 4.11 (q, J = 7.0 Hz, 2H),
1.30 (t, J = 7.0 Hz, 3H).
*
Cl
Cl
O
14c
14f
N
*
*
Racemate.
Table 5
Selectivity of key compounds
O
2D ROESY
H
N
N
correlation
GPR310
pIC50
GPR610
pIC50
GPR1210
pIC50
H
Compound
N
N
9h
9k
14d
AF64394
14e
6.6
6.7
7.0
7.3
6.8
5.1
4.9
5.1
5.1
4.9
5.3
5.1
4.8
4.9
5.2
N
Ph
8b
16. Prepared according to Scheme 1 from 3-aminopyrazole.
17. AF64394:
7-Chloro-5-phenyl-[1,2,4]triazolo[1,5-a]pyrimidine11
(100 mg,
0.433 mmol) and 4-chloro-2-isopropoxybenzylamine (114 mg, 0.650 mmol)
in EtOH (20 mL) were stirred overnight at room temperature. Concentration in
vacuo and purification with preparative HPLC afforded N-(4-chloro-2-isoprop-
compounds are only of modest potency and are relatively lipo-
philic (e.g., AF34394 clogP = 5.2, LLE19 = 2.1, and 14b clog P 3.1
and LLE 3.2) the series shows promising SAR and represents an
excellent start point for further optimisation.
oxybenzyl)-5-phenyl-[1,2,4]triazolo[1,5-a]pyrimidin-7-amine
(AF64394)
d 1.23 (d,
(80 mg, 47%) as
a
white solid. 1H NMR (400 MHz, DMSO-d6)
J = 6.02 Hz, 6H), 4.65–4.76 (m, 3H), 6.87 (s, 1H), 6.92 (dd, J = 8.16, 1.88 Hz, 1H),
7.10 (d, J = 2.01 Hz, 1H), 7.33 (d, J = 8.03 Hz, 1H), 7.45–7.51 (m, 3H), 8.07–8.13
(m, 2H), 8.51 (s, 1H), 8.63 (br s, 1H).
Supplementary data
18. See Supplementary information.
19. Abad-Zapatero, C.; Metz, J. Drug Discovery Today 2005, 10, 464.
Supplementary data associated with this article can be found, in
the online version, at http://dx.doi.org/10.1016/j.bmcl.2014.09.077.