Synthesis and binding profile of haloperidol-based bivalent ligands targeting dopamine D2-like receptors

Article abstract of DOI:10.1016/j.bmcl.2014.06.079

Homodimers of dopamine D₂-like receptors are suggested to be of particular importance in the pathophysiology of schizophrenia and, thus, serve as promising targets for the discovery of atypical antipsychotics. This study describes the development of a series of novel bivalent molecules with a pharmacophore derived from the dopamine receptor antagonist haloperidol. These dimers were investigated in comparison to their monomeric analogues for their D_2long, D_2short, D₃, and D₄ receptor binding and the ability to bridge two neighboring receptor protomers. Radioligand binding studies provided diagnostic insights when Hill slopes close to two for the bivalent ligand 13 incorporating 22 spacer atoms and a comparative analysis with monovalent control ligands indicated a bivalent binding mode with a simultaneous occupancy of two neighboring binding sites.

Full text of DOI:10.1016/j.bmcl.2014.06.079

Bioorganic & Medicinal Chemistry Letters 24 (2014) 3753–3756
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
Synthesis and binding profile of haloperidol-based bivalent ligands
targeting dopamine D2-like receptors
Ismail Salama a,b, Stefan Löber a, Harald Hübner a, Peter Gmeiner a,
⇑
a Department of Medicinal Chemistry, Emil Fischer Center, Friedrich-Alexander University, Schuhstraße 19, D-91052 Erlangen, Germany
b Department of Pharmaceutical Chemistry, Faculty of Pharmacy, Suez Canal University, 41522 Ismailia, Egypt
a r t i c l e i n f o
a b s t r a c t
Article history:
Homodimers of dopamine D2-like receptors are suggested to be of particular importance in the patho-
physiology of schizophrenia and, thus, serve as promising targets for the discovery of atypical antipsy-
Received 14 May 2014
Revised 24 June 2014
Accepted 26 June 2014
Available online 3 July 2014
chotics. This study describes the development of
a series of novel bivalent molecules with a
pharmacophore derived from the dopamine receptor antagonist haloperidol. These dimers were investi-
gated in comparison to their monomeric analogues for their D2long, D2short, D3, and D4 receptor binding
and the ability to bridge two neighboring receptor protomers. Radioligand binding studies provided diag-
nostic insights when Hill slopes close to two for the bivalent ligand 13 incorporating 22 spacer atoms and
a comparative analysis with monovalent control ligands indicated a bivalent binding mode with a simul-
taneous occupancy of two neighboring binding sites.
Keywords:
Bivalent ligands
Dopamine
Binding affinity
GPCR dimers
D2 receptor
D3 receptor
D4 receptor
Ó 2014 Elsevier Ltd. All rights reserved.
G-protein-coupled receptors (GPCRs) are transmembrane
receptors that mediate most of their intracellular actions
through pathways involving an activation of G-proteins.1 Because
GPCR-promoted signaling is involved in a great number of patho-
physiological processes, members of this large family of proteins
are targeted most frequently at the development of therapeutics
for numerous diseases.2,3
Employing the antipsychotic drugs haloperidol and haloperidol
decanoate as lead compounds,20 we envisioned to synthesize and
biologically investigate dimeric derivatives thereof. Thus, we
designed a prototypical structure of bivalent molecules using poly-
methylene moieties or triazolyl-linked spacer elements, which
were bound via carboxylate functions to the tertiary alcohol group
of the pharmacophore unit. Subsequent radioligand binding stud-
ies indicated a bivalent binding mode with a simultaneous occu-
pancy of two adjacent binding sites when Hill slopes close to two
were observed for the test compound 13 comprising a 22-atom
spacer length.
A selection of spacer groups have been reported in the litera-
ture, varying from polymethylene, polyamide, polyethyleneglycol,
polyproline, and piperazine containing moieties, all with differing
degrees of flexibility and lipophilicity. Designing a collection of
homobivalent dopamine receptor ligands, we incorporated either
methylene or ethylene glycol units with or without triazole or
piperazine rings to build spacers with a length varying from 4 to
26 atoms.
Several lines of evidence suggest that some GPCRs including
5,6
7
D1,4 D2 and D3 receptors may exist in dimeric or oligomeric
form modulating ligand pharmacology, signal transduction and
cellular trafficking.8 Using functional complementation assays,
communication between protomers of the D2 receptors could be
observed leading to a significant modulation of GPCR activation
in transfected cell lines.9 According to very recent findings,10
homodimers of dopamine D2-like receptors might be of particular
importance in the pathophysiology of schizophrenia serving as
promising targets for the discovery of atypical antipsychotics.
Bivalent ligands11–14 were developed consisting of two pharma-
cophore moieties connected by a spacer with an appropriate length
allowing the two head groups to interact with two orthosteric
binding sites of adjacent receptors of a dimer or with another rel-
evant allosteric site at the same protomer.15–19
For comparative purposes, monovalent fragments containing
one-half of the linker were synthesized for selected compounds.
The bivalent ligands 1–13 were synthesized by treatment of halo-
peridol with dicarboxylic acid chlorides of varying length in pres-
ence of DIPEA to give the bis-haloperidol derivatives bearing
ester functionalities at their point of attachment (Scheme 1).
Analogously, the respective monovalent compounds 14–17 were
⇑
Corresponding author. Tel.: +49 9131 8524116.
E-mail address: peter.gmeiner@fau.de (P. Gmeiner).
http://dx.doi.org/10.1016/j.bmcl.2014.06.079
0960-894X/Ó 2014 Elsevier Ltd. All rights reserved.

3754
I. Salama et al. / Bioorg. Med. Chem. Lett. 24 (2014) 3753–3756
Cl
Cl
O
O
O
O
O
O
N
N
( )3
( )3
( )n
F
N
N
F
O
O
R'
O
R''
N
Cl
( )3
bivalent ligands
F
O
O
1
5
9
: n= 10
: n= 0
: n= 4
2: n= 1
3: n= 2
4: n= 3
6: n= 5 10: n= 12
7: n= 6 11: n= 14
8: n= 8 12: n= 16
O
N
Cl
( )3
13
: n= 18
19
20
: R', R''= -CH2CH2-
: R', R''= CH3
F
a
b
b
O
O
OH
O
a
O
N
Cl
N
Cl
( )3
18
( )3
haloperidol
F
F
a
O
O
monovalent ligands
O
N
O
R
O
O
N
Cl
N
Cl
( )3
( )3
14
21
: R= p-methylphenyl
F
F
15: R= methyl
16: R= ethyl
17: R= n-nonyl
Scheme 1. Reagents and conditions: (a) corresponding acid chloride or diacid chloride, DIPEA, CH2Cl2, rt, 10 min–24 h, 59–70%; (b) corresponding amine or diamine, FeCl3,
CH2Cl2, rt, 48–72 h, 43–64%.
prepared by treatment of haloperidol with carboxylic acid chlo-
rides of varying length in presence of DIPEA. The test compounds
19–21 were prepared by reaction of the central building block 18
with the corresponding amine or diamine under Michael reaction
conditions.
to displace the radiolabeled antagonist [3H]spiperone from human
D2long, D2short, D3, and D4.4 receptors stably expressed in Chinese
hamster ovary cells (CHO). D1 receptor binding affinities were
determined utilizing striatal membranes and the D1 selective radi-
oligand [3H]SCH 23390.
The synthesis of bivalent ligands featuring a bis-triazol linker
unit started from 4-pentynoic acid, which was converted to
the corresponding acid chloride upon treatment with oxalyl
chloride (Scheme 2). Subsequent reaction with haloperidol affor-
ded the respective ester 22, which underwent copper-catalyzed
azide–alkyne cycloaddition (CuAAC) with 1,8-diazidooctane and
1,10-diazidodecane in presence of cuprous tetraacetonitrile hexa-
fluorophosphate as a catalyst yielding the bistriazoles 23 and 24,
respectively. Reaction of 22 with diazidotriethyleneglycol and
diazidotetraethylene glycol yielded the analogous derivatives 25
and 26, respectively.
To demonstrate the importance of the basic nitrogen of the
piperidine nucleus of haloperidol, we synthesized the non-basic
control compound 27 featuring an amide moiety instead of the
tertiary amine. Thus, 27 was derived by amide coupling of
4-(4-fluorophenyl)-4-oxobutanoic acid chloride with 4-hydroxy-
4-(4-chlorophenyl)piperidine in the presence of Hünig’s base.
To assess the receptor binding properties, the mono- and
dimeric haloperidol derivatives were investigated for their ability
Our initial investigations were directed to careful comparison of
the dimeric ligand 1 and 3 with their monovalent analogues 15 and
16. The test compounds showed significant specific binding for
D2long, D2short, D3, and D4 and clearly reduced D1 affinity (Table 1)
resulting in Ki values in the single digit and double digit nanomolar
range with slightly higher D3 affinities for the dimeric test com-
pounds. It is noteworthy that the D3/D4 selectivity is reversed
between monomers and dimers. Whereas 1 and 3 displayed a 2–
4-fold preference for D3 over D4, approximately 8–15-fold D4/D3
selectivity has been observed for 15 and 16.
To gain diagnostic insights into the binding modes of our biva-
lent ligands, we analysed the profiles of the competition binding
curve in detail. Competitive one-site binding curves that follow
the law of mass action have a slope of one. Antagonists tend to
have Hill slope (nH) values close to one because they fail to differ-
entiate between inactive, uncoupled receptor and active state of
the GPCR, thereby following the general law of mass action for sin-
gle site competition.21 Examination of the binding curve of the
monomeric ligands 15 and 16 revealed Hill slopes close to one

I. Salama et al. / Bioorg. Med. Chem. Lett. 24 (2014) 3753–3756
3755
O
O
COOH
O
a
N
Cl
22
F
b
X
N
N
N
N
N
N
O
O
O
O
O
O
N
Cl
Cl
N
23
24
: X= -(CH2)6- 25
: X= -(CH2OCH2)3-
: X= -(CH2OCH2)4-
F
F
F
: X= -(CH2)8- 26
O
OH
X
O
d
N
Cl
O
O
X= OH
X= Cl
F
c
27
Scheme 2. Reagents and conditions: (a) (i) oxalyl chloride, CH2CI2, rt, 28 h; (ii) haloperidol, DIPEA, CH2CI2, rt, 24 h, 47%; (b) diazide, Cu(CH3CN)4PF6, 40 °C, 4 h, 38–74%; (c)
oxalyl chloride, CH2CI2, rt, 21 h; (d) 4-hydroxy-4-(4-chlorophenyl)piperidine, DIPEA, CH2CI2, 0 °C to rt, 22 h, 47%.
Table 1
Receptor binding data for the bivalent compounds and their monomeric counterparts employing porcine dopamine D1 receptor and the human subtypes D2long, D2short, D3, and
D4.4
Ki valuesa (nM) SEM and Hill slopesb,c [absolute value of nH SEM]
[3H]SCH23390
pD1
[3H]spiperone
Compd
n⁄d
hD2long
hD2short
hD3
hD4.4
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
19
20
21
23
24
25
26
4
5
6
7
8
130 42c
220 26
220 27
6.7 2.5 [1.2 0.17]
12 3.8 [1.4 0.16]
5.4 2.1 [1.4 0.11]
12 4.0 [1.5 0.19]
14 9.5 [1.7 0.35]c
24 6.5 [0.77 0.04]
31 8.1 [1.5 0.22]
11 3.3 [1.3 0.15]
11 2.3 [1.7 0.14]
6.4 1.7 [0.84 0.04]
9.1 1.7 [0.73 0.055]
39 12 [0.67 0.07]
210 94 [1.3 0.18]
160 14 [1.0 0.073]
220 35 [1.0 0.071]c
380 31 [1.4 0.24]
82 3.6 [1.3 0.13]
500 36 [1.8 0.13]
270 26 [0.78 0.025]
56 19 [1.3 0.1]
33 4.5 [1.2 0.025]c
41 7.8 [0.85 0.0]c
360 14 [0.95 0.15]c
300 79 [1.4 0.4]c
3.0 0.43 [1.2 0.15]
3.3 0.79 [1.2 0.19]
2.5 0.51 [0.7 0.042]
5.6 0.94 [0.66 0.059]
15 2.6 [0.79 0.085]
66 29 [0.85 0.11]
120 13 [1.0 0.043]
140 21 [1.3 0.18]c
160 6.5 [1.0 0.075]
25 2.1 [1.1 0.11]
2.3 0.67 [0.92 0.051]
2.6 0.65 [1.2 0.18]
2.9 0.7 [0.67 0.047]
6.5 2.3 [0.56 0.019]
31 11 [0.71 0.062]
46 13 [0.98 0.082]
62 10 [0.89 0.16]
280 21 [1.2 0.071]c
150 9.1 [0.99 0.013]
27 4.5 [1.0 0.06]
280 63
920 310
1200 410
1400 250
1100 440
1400 420c
6800 280c
930 240c
1400 71c
1500 86
550 240c
59 1.5
9
520 130 [0.98 0.18]
490 140 [1.7 0.05]
2200 210 [0.75 0.071]
490 48 [1.4 0.2]
10
12
14
16
18
20
22
1400 420 [1.1 0.035]c
570 31 [1.5 0.3]
160 19 [1.1 0.14]
850 120 [1.7 0.076]
130 26 [1.6 0.075]
9.3 4.2 [1.3 0.29]
2.5 0.1 [1.0 0.05]c
5.8 1.9 [0.68 0.11]c
33 4.7 [0.92 0.088]
29 2.1 [1.0 0.035]c
8.7 4.7 [0.88 0.18]c
19 8.5 [0.93 0.035]c
89 16 [0.85 0.071]c
41 5.7 [1.6 0.35]c
46 9.2 [0.9 0.071]c
18000 7800 [1.1 0.35]c
5.9 0.98 [1.1 0.1]
310 23 [1.8 0.1]
56 24 [1.1 0.24]
280 25 [1.6 0.1]
47 18 [1.2 0.14]
4.5 0.82 [1.1 0.2]
4.5 1.2 [1.1 0.19]
1.6 1.3 [0.85 0.35]c
0.46 0.11 [0.88 0.035]c
1.1 0.39 [0.7 0.11]
2.2 1.6 [0.9 0.0]c
72 5.0c
3.9 0 [0.8 0.21]c
41 11 [1.1 0.14]c
3.5 1.6 [1.1 0.35]c
9.5 1.1 [0.96 0.031]
47 0.71 [1.6 0.64]c
30 3.5 [0.95 0.35]c
1.8 0.78 [1.2 0.21]c
5.9 1.9 [0.9 0.28]c
9.8 4.6 [1.4 0.32]c
19 11 [1.1 0.14]c
1700 490 [0.8 0.21]c
2.6 0.33 [1.0 0.046]
140 59c
360 120
260 21c
200 49c
49 30c
1.8 0.57 [0.73 0.035]c
2.5 0.61 [0.73 0.1]
3.2 0.21 [0.75 0.0]c
1.9 0.71 [0.93 0.035]c
1.3 0.21 [0.88 0.25]c
4.9 1.8 [0.63 0.11]c
1.0 0.68 [1.0 0.071]c
1.6 1.5 [0.8 0.071]c
860 91 [0.85 0.14]c
0.81 0.14 [1.0 0.056]
10
10
1.8 1.1 [0.73 0.32]c
1.6 0.50 [0.9 0.14]c
5.4 0.21 [0.63 0.035]c
3.0 2.3 [1.4 0.21]c
3.8 2.3 [1.0 0.11]c
470 250 [0.7 0.0]c
0.85 0.18 [1.1 0.097]
20
22
23
26
86 15c
120 7.1c
150 7.1c
19000 1400c
9.5 0.91
27
haloperidol
a
b
c
Ki values SEM derived from 4–10 experiments each done in triplicate.
Hill slopes SEM are derived from the same binding curves recorded for the determination of Ki values; the original nH was negative but id displayed as absolute value.
Ki values SD and Hill slopes [absolute value of nH] SD derived from two experiments each done in triplicate.
Overall number of spacer atoms.
d

3756
I. Salama et al. / Bioorg. Med. Chem. Lett. 24 (2014) 3753–3756
(nH = 0.8–1.3) indicating antagonist properties and a binding mode
reflecting that one receptor protomer binds one ligand. Interest-
ingly, careful analysis of the heterologous competition experi-
ments that we conducted with the dimers 1 and 2 revealed a
steepening of the competition curves leading to Hill slopes of
1.2–1.7, which might indicate positive cooperative binding to a
receptor dimer and, thus, displacement of two equivalents of the
radioligand. Positive cooperativity is usually observed as a conse-
quence of an allosteric effect inducing conformational crosstalk
within one receptor protomer or modulation of the interaction
between two protomers of a receptor dimer.22,23 The highest Hill
slope of 1.8 was calculated for the interaction of D2long and D3 with
the bivalent ligand 13, incorporating a 22 atom spacer. To learn
whether this effect depends on the length of the spacer arm, the
dimeric homologues 1–12 with 4–20 atom spacers, were investi-
gated. The competitive binding studies gave Hill coefficients
between 0.66 and 1.4 for D2long and between 0.67 and 1.7 for D3
indicating that the distance below 22 atoms leads to a reduction
of a bridging binding mode. To conclude the effect of the nature
and physicochemical properties of the linker, the dimers with
nitrogen atoms in the linker and those with hydrophilic oligoethyl-
ene glycol units were investigated. We observed an increase in the
binding affinity for D2, D3 and D4 compared to the non-nitrogen
containing analogues with Hill slopes close to one for 19, 23, 24
antagonist haloperidol. At the critical spacer length of 22 atoms,
a bivalent binding mode has been suggested with a simultaneous
occupancy of two neighboring binding sites. The binding proper-
ties of the dimeric ligands depended on the structure and the
length of the spacer arm. To improve the ligand affinity of the biva-
lent ligand 13, further structural evolution is envisioned.
Acknowledgment
Financial support from the German Research Foundation (DFG
Gm 13-8) is acknowledged.
Supplementary data
Supplementary data associated with this article can be found, in
the online version, at http://dx.doi.org/10.1016/j.bmcl.2014.
06.079.
References and notes
1. Gether, Ulrik; Asmar, Fazila; Meinild, Anne Kristine; Rasmussen, Soren G. F. J.
Pharmacol. Toxicol. 2002, 91, 304.
2. Overington, J. P.; Al-Lazikani, B.; Hopkins, A. L. Nat. Rev. Drug Disc. 2006, 5, 993.
3. Lagerstrom, M. C.; Schioth, H. B. Nat. Rev. Drug Disc. 2008, 7, 339.
4. Rashid, A. J.; So, C. H.; Kong, M. M.; Furtak, T.; El-Ghundi, M.; Cheng, R.;
O’Dowd, B. F.; George, S. R. Proc. Natl. Acad. Sci. U.S.A. 2007, 104, 654.
5. Ng, G. Y. K.; O’Dowd, B. F.; Lee, S. P.; Chung, H. T.; Brann, M. R.; Seeman, P.;
George, S. R. Biochem. Biophys. Res. Commun. 1996, 227, 200.
6. Zawarynski, P.; Tallerico, T.; Seeman, P.; Lee, S. P.; O’Dowd, B. F.; George, S. R.
FEBS Lett. 1998, 441, 383.
and 26 while for 25 Hill slopes between 1.4 and 1.6 for D2short
,
D3 and D4 were calculated. Thus, the length, nature and physico-
chemical properties of the spacer clearly affect the binding mode
and binding affinity of the bivalent ligands.
7. Nimchinsky, E. A.; Hof, P. R.; Janssen, W. G. M.; Morrison, J. H.; Schmauss, C. J.
Biol. Chem. 1997, 272, 29229.
Overall, although most Hill slopes are close to unity, a wide
range of values (0.56–1.8) have been observed for some ligands.
This reflects multiple binding modes of the newly designed halo-
peridol derivatives. Hill slopes smaller than one indicate a biphasic
curve resulting from coexistence of high affinity and low affinity
binding sites. Typically, agonists can cause such biphasic curve
which is due to their ability to bind G protein-coupled, active
receptors with a higher affinity than uncoupled, inactive receptors.
Because our ligands are based on an antagonist pharmacophore,
small Hill slopes of this study are suggested to indicate a mixture
of monomeric receptors and receptor dimers consisting of cooper-
ating protomers. If the binding affinity of a ligand to a binding
pocket of the receptor dimer is different to the affinity to the
monomer, a biphasic curve and Hill slopes smaller than one will
be observed. Hill slopes larger than unity indicate positive cooper-
ative effects between receptor protomers resulting from allosteric
modulation by receptor–receptor interaction or the existence of a
receptor—bridging binding mode leading to the displacement of
two equivalents of radioligand. The latter property has been
ascribed to the bivalent ligand 13, because Hill slopes close to
two were observed for all four subtypes of the D2-like receptors.
Moreover, this effect strongly depends on the length of the linker.
In summary, we have synthesized a number of novel monova-
lent and homobivalent ligands based on the dopamine receptor
8. Milligan, G. Biochim. Biophys. Acta 2007, 1768, 825.
9. Han, Y.; Moreira, I. S.; Urizar, E.; Weinstein, H.; Javitch, J. A. Nat. Chem. Biol.
2009, 5, 688.
10. Wang, M.; Pei, L.; Fletcher, P. J.; Kapur, S.; Seeman, P.; Liu, F. Mol. Brain 2010, 3,
25.
11. Portoghese, P. S. J. Med. Chem. 2001, 44, 2259.
12. Portoghese, P. S.; Larson, D. L.; Sayre, L. M.; Yim, C. B.; Ronsisvalle, G.; Tam, S.
W.; Takemori, A. E. J. Med. Chem. 1855, 1986, 29.
13. Zhang, Y.; Gilliam, A.; Maitra, R.; Damaj, M. I.; Tajuba, J. M.; Seltzman, H. H.;
Thomas, B. F. J. Med. Chem. 2010, 53, 7048.
14. Bhushan, R. G.; Sharma, S. K.; Xie, Z.; Daniels, D. J.; Portoghese, P. S. A. J. Med.
Chem. 2004, 47, 2969.
15. Kuhhorn, J.; Hubner, H.; Gmeiner, P. J. Med. Chem. 2011, 54, 4896.
16. Kuhhorn, J.; Gotz, A.; Hubner, H.; Thompson, D.; Whistler, J.; Gmeiner, P. J. Med.
Chem. 2011, 54, 7911.
17. Huber, D.; Lober, S.; Hubner, H.; Gmeiner, P. Bioorg. Med. Chem. 2012, 20, 455.
18. McRobb, F. M.; Crosby, I. T.; Yuriev, E.; Lane, J. R.; Capuano, B. J. Med. Chem.
2012, 55, 1622.
19. Birnkammer, T.; Spickenreither, A.; Brunskole, I.; Lopuch, M.; Kagermeier, N.;
Bernhardt, G.; Dove, S.; Seifert, R.; Elz, S.; Buschauer, A. J. Med. Chem. 2012, 55,
1147.
20. Löber, S.; Hübner, H.; Tschammer, N.; Gmeiner, P. Trends Pharmacol. Sci. 2011,
32, 148.
21. Urban, J. D.; Clarke, W. P.; von Zastrow, M.; Nichols, D. E.; Kobilka, B.;
Weinstein, H.; Javitch, J. A.; Roth, B. L.; Christopoulos, A.; Sexton, P. M.; Miller,
K. J.; Spedding, M.; Mailman, R. B. J. Pharmacol. Exp. Ther. 2007, 320, 1.
22. Smith, N. J.; Milligan, G. Pharmacol. Rev. 2010, 62, 701.
23. Albizu, L.; Balestre, M. N.; Breton, C.; Pin, J. P.; Manning, M.; Mouillac, B.;
Barberis, C.; Durroux, T. Mol. Pharmacol. 2006, 70, 1783.