5400
N. Nebel et al. / Bioorg. Med. Chem. Lett. 24 (2014) 5399–5403
H3CO
H3CO
N
O
O
N
H
N
N
N
H
N
S
BP 897
FAUC346
18F
O
O
O
11CH2CH2CH3
N
N
N
HO
N
H
O
S
[
11C]-(+)-PHNO
[
18F]LS-3-134
H3CO
N
Cl
Cl
O
N
N
H
N
N
N
N
N
H
F
F
N
3 (ref. 30)
4 (ref. 28)
Figure 1. Structures of D3 ligand lead compounds.
18F-labeled lead series derivatives of the pyridinylcarboxamide
318,30 and phenylazocarboxamide 428 (Fig. 1). These newly
developed 18F-labeled tracers needed further optimization, since
interfering binding to 5-HT1A receptors was observed and the
lipophilicity of these candidates needed to be improved.
Compound 3 showed a high octanol/water partition coefficient
(logP) of 4.39, which is predictive of permeability to the blood–
brain barrier (BBB). However, logP values in the range 2–3.5 are
considered optimal; within a class of structurally related com-
pounds, higher logP values generally impart higher non-specific
binding, and less signal-to-noise in PET recordings.31 This is partic-
ularly an issue for detecting receptors of rather low abundance,
such as the D3 site. The calculated logP of the compound 4 was
similar (4.55) (Fig. 1).
Previous studies by other groups have shown that hydroxyl-
ation of the alkyl chain lowers the logP value, whilst retaining high
affinity, and in some cases even increasing D3 selectivity.15 Thus,
we predicted that the lipophilicity of lead compounds 3 and 4
should be significantly reduced by the introduction of a C-3 hydro-
xyl group. The aspect of further SAR studies with lead compound 4
despite its high logP value is the observation that the aromatic
nucleus of phenylazocarboxylic esters is highly activated for nucle-
ophilic aromatic substitution, and subsequent reactions based at
the carbonyl group are also easily obtained. These are favorable
attributes for very efficient 18F-labeling.28,32,33
We now report further SAR developments on pyridinylcarboxa-
mide 3 and phenylazocarboxamide 4 with the intention to generate
a high affinity and highly selective D3 receptor ligand. In addition,
the lipophilicity of a series of compounds was reduced by introduc-
ing a hydroxyl group at the butyl spacer in both classes of lead
compounds. Moreover, we varied the substituents at the aromatic
ring of our new series of compounds for three reasons: First, our
previous studies on the influence of the 2,3-dichloro substituent34
and the 2-chloro or mixed chloro and methoxy substituents17 at
the phenylpiperazinyl moiety of structurally related derivatives
revealed a significant influence on D3 affinity and beneficial D3
subtype selectivity. Second, the introduction of a methoxy substitu-
ent could be most suitable to further improve the hydrophilicity of
candidate ligands,35,36 and third, the introduction of the para-cyano
substituent was envisaged, since this substitution pattern has been
previously reported for a D3 antagonist ligand.37
the new derivatives is outlined in Scheme 1 and followed previ-
ously published synthetic routes.12,15,18,28,38
The primary amines 8a, 8c, 8e and 8g were obtained starting
from the commercially available substituted phenylpiperazine by
N-alkylation with 4-bromobutylphthalimide (6a) and subsequent
hydrazinolysis in yields of 53–66% (Scheme 1).12 Compound 9d
was synthesized by coupling of 6-fluoropyridine-3-carbonyl chlo-
ride (1) with the aminobutyl-substituted phenylpiperazine 8g in a
yield of 45%.18 The synthesis of the hydroxylated pyridinylcarboxa-
mides 9a, 9b, 9c and 9e and phenylazocarboxamides 10b, 10e, 10g
and 10h are also illustrated in Scheme 1. 2-(2-Bromoethyl)oxirane,
which is synthesized from commercially available 4-bromo-
1-butene and mCPBA, was reacted with potassium phthalimide to
form the 2-(2-(oxirane-2-yl)-ethyl)isoindoline-1,3-dione 6b.15,39
The conformationally strained epoxides react cleanly with amines
5a, 5b, 5d–f to yield amino alcohols 7b, 7d, 7f, 7h and 7i when
the ring opening occurs in a regioselective manner at the least
substituted side of the oxirane. This is the case for all of the phenyl-
piperazines, leading to racemic mixtures of products in high yields
(86–93%). The hydroxybutylamines 8b, 8d, 8f, 8h, 8i were obtained
after treatment of 7b, 7d, 7f, 7h, 7i with hydrazine in markedly
varying yields (36–70%).15 In the final coupling step, 8b, 8f, 8h, 8i,
as well as unhydroxylated compounds 8a, 8c, 8e and 8g were then
reacted with tert-butyl 2-(4-fluorophenyl)azocarboxylate (2) in the
presence of K2CO3 in an nucleophilic substitution to afford 10a–h in
low yields (10–45%).28 The synthesis of the respective desired
hydroxylated pyridinylcarboxamide derivatives 9a–c and 9e was
accomplished by N-acylation of 8b, 8d, 8f and 8h with 6-fluoropyr-
idine-3-carbonyl chloride 6a in yields of 43–64%.15,18
The lipophilicity calculations of the compounds 9a–e and 10a–h
were performed using the software ChemDraw Ultra (Cambridge-
Soft, Perkin Elmer). The range of calculated logP (clogP) values of
the compounds 9a–e and 10a–h were between 2.05 and 6.18
(Table 1). The introduction of the hydroxyl group at the butyl
spacer (R4) of lead compound 4 (clogP 4.55) lowered the logP
value by 0.82 units as demonstrated by the clogP of compound
10e (clogP 3.73, Table 1). The same effect was found by comparing
compounds 10a with 10b, and 10f with 10g (decreased clogP by
0.86 units). In the case of structures of class 9, the difference in
clogP units induced by the hydroxyl group was 0.63, as seen in
the comparisons of lead compound 3 with 9b, and 9d with 9e.
Thus, introduction of a hydroxyl function in the alkyl linking chain
consistently lowers lipophilicity. Varying the substitution pattern
We demonstrate the synthesis of several new fluorinated D3
receptor ligands with potential as PET tracers. The synthesis of
Nebel, Natascha
