F. Micheli et al. / Bioorg. Med. Chem. Lett. 26 (2016) 1329–1332
1331
OH
O
OH
Cl
H
N
NH2
Br
R
R
iv
R
i; ii; iii
O
O
O
O
NH
NH
R
R
vi
v
Scheme 1. Reagents and conditions: (i) HMTA, CHCl3, 16 h, rt; (ii) 37% HCl, EtOH,
12 h, rt; (iii) NaBH4, MeOH, 30 min, 0 °C; (iv) chloroacetylchloride, TEA, DCM, 1 h,
0 °C; (v) t-BuOK, THF, 1 h, 0 °C—rt; (vi) LiAlH4, THF, 1 h, reflux.
Cl
N
O
N
N
N
Cl
Br
HS
S
HO
N
N
HET
HET
HET
vii; viii
ix
O
NH
R
O
N
N
N
N
S
x
R
HET
Scheme 2. Reagents and conditions: (vii) T3P in AcOEt, 4-methyl-3-thiosemicar-
bazide, DIPEA, DMF, 12 h, rt; (viii) 4 M NaOH, 40 min, reflux; (ix) K2CO3, methanol/
acetone, 12 h, rt; (x) Na2CO3, NaI, DMF, 12 h, 60 °C.
selectivity of this new class versus the DA D1 and D4 receptors, and
Muscarinic M1 and M3 receptors.
All the compounds with pKi > 7.0 were tested to evaluate their
functional activity. No sign of agonism at the DA D3 receptor was
observed (fpKi < 5 for all the compounds tested) while the antago-
nistic fpKi values are reported in Tables 1–3.
Selected compounds went through generic developability
screens such as CYPEX bactosome P450 inhibition and rat and
human in vitro clearance in liver microsomes early in the screening
cascade to predict their PK and developability profiles.
The first compound identified during this ‘scaffold hopping’
exercise was the racemic derivative 6, which showed micromolar
affinity at the DA D3 receptor. The mixture was therefore separated
by chiral HPLC and the most potent enantiomer (7) was evaluated.
In terms of selectivity, this compound was completely inactive at
the DA D1, DA D4 and at the M3 receptors, showing functional
pKi (fpKi) = 5.5 at the M1 receptor. A 10-fold difference with hERG
electrophysiological measure was also observed.
Figure 3. Derivative 26 (orange) fitted in the DA D3 receptor model together with 2
(atom type). The interaction between the basic nitrogen of the morpholine or [3.1.0]
hexane moieties and Asp3.32 is clearly visible (red dotted line).
active enantiomer 9. This substitution led to approximately 10-fold
increased affinity at the DA D3 receptor, 100-fold selectivity over
the DA D2 receptor, and more than 50-fold selectivity over the
hERG channel. This substitution had no effect on the affinity at
DA D1 and D4 receptors (pKi < 4.0), but resulted in a slight increase
in affinity at the M1 and M3 receptors (fpKi = 6). Increased inhibi-
tion of CYP P450 was observed with this compound, which showed
IC50 values greater than 6
lM on all isoforms tested. A slight
increase in the in vitro clearance in human (hCli = 40.0
l
L/min/
mg protein) with no observed changes in rat microsomes. This
compound also showed Fa% = 63%, Vss = 2.5 (l/kg), and Clb = 13.5
(mL/min/kg) leading to T1/2 = 3.5 (h); Eh was low (0.1) leading to
58% bioavailability. The percentage of fraction unbound in blood
and plasma was definitely lower with blood Fu% equal to 6.2%
and brain Fu% equal to 7.1%. This decrease might be linked to the
increased overall lipophilicity (cLogP from 2.13 to 2.87)14 due to
the introduction of the –CF3 moiety.
Increasing the length of the linker portion from 3 to 4 carbon
atoms (racemic derivative 10 and its enantiomer 11) had no major
impact on the primary pharmacological profile.
The introduction of an additional fluorine atom on the aromatic
ring linked to the morpholine moiety (12) led to slightly reduced
affinity at the DA D3 receptor. Replacement of the –CF3 with a
methyl group (13) resulted in decreased affinity at the DA D3
receptor, while a 4-bromine alone (14) was almost identical to
the 2-F, 4-CF3 substitution.
The next steps of the exploration focused on the heteroaromatic
portion of the scaffold, where the methyl oxazole was initially
replaced by a thiophene moiety (15) with similar affinity to 8.
The introduction of a nitrogen atom (thiazole-2-yl, 16) resulted
in decreased affinity at the DA D3 receptor, which was even worse
with the methyl-pyrazol-5-yl derivative 17. In contrast, a single
nitrogen atom (methyl-pyrrol-2-yl) 18 was very similar to the
original derivative 8.
From a developability point of view, compound 7 showed IC50
values greater than 30
CYP1A2, CYP2C9, CYP2C19, CYP2D6, CYP3A4 DBOMF and CYP3A4
7BQ), low-moderate clearance in human (hCli = 21.7 L/min/mg
L/min/mg protein)
lM on all CYP P450 isoforms tested (namely
l
protein), and low clearance in rat (rCli) (<9
microsomes.
l
In order to fully assess the potential of the scaffold, its proper-
ties were further explored in vivo11 in a portal vein-cannulated rat
model.12 Compound 7 had 100% fraction absorbed (Fa%), low distri-
bution volume (Vss = 2.6 l/kg), and moderate clearance as expected
from the in vitro parameters (Clb = 40.0 mL/min/kg) leading to a
relatively short half-life (T1/2 = 0.8 h). Nonetheless, hepatic extrac-
tion (Eh) was low (Eh = 0.1) resulting in almost full 100% bioavail-
ability. The percentage of fraction unbound (Fu%) in blood and
plasma was also measured with 17% of the compound in free form
in blood; in the brain this compound was present as ‘free’ for more
than 50%.
The importance of ‘developability’ and ‘drug efficiency’,13 led
quite naturally to the next exploration step that focused on replac-
ing the fluorine atom present on the aromatic ring of compound 6
with a trifluoromethyl group, resulting in the racemate 8 and in the