C. Mugnaini, et al.
Bioorganic&MedicinalChemistryLettersxxx(xxxx)xxxx
Scheme 3 describes the synthesis of compounds of class C. Com-
pounds 12a and 12b were obtained reacting ethyl 3-phenyl-3-ox-
opropanoate with thiourea (11a) or N-methylthiourea (11b), respec-
tively, while cyclization of 11b with ethyl 3-(4-tert-butylphenyl)-3-
oxopropanoate (obtained by acylation of potassium salt of ethyl mal-
onate with 4-(tert-butyl)benzoyl chloride) afforded 12c. The final 2-
phenylthiopyrimidinones 13a-e were prepared starting from 12a-c by a
Ulmann type coupling reaction using the appropriate phenylboronic
acid and microwave heating at 85 °C.
HN
N
HO
NO2
NH
CF3
O
CF3
NH
H3CS
N
O
N
N
rac-BHFF
GS39783
BHF177
Finally, compounds of class D were prepared from commercially
available anilines 14a-c (Scheme 4) by N-acylation with the appro-
reduction to afford the alkylated anilines 16a-f.
Fig. 1. Chemical structure of selected GABABR PAMs.
dimethylpyrazole, a motif which is largely represented in medicinal
chemistry. Compounds of classes B and C can be considered an evolu-
tion of BHF177 of which they maintain the core structure, with the
introduction of groups potentially able to further interact with the
target through H-bond formation and lipophilic interactions. Finally,
compounds of class D draw inspiration from rac-BHFF, a molecule for
which concern about chemical stability and developability can be
raised for at least three points: an O-aryl γ-lactone group and a chiral
carbon atom bearing a notably acidic tertiary alcohol. Here, the ben-
zofurane scaffold is replaced by a 2-quinolone system, which allows to
obtain achiral, more stable and handy compounds, while retaining to
some extent the same substitution pattern.
Cyclization with differently substituted ethyl malonates provided
the 4-hydroxyquinolin-2(1H)-one derivatives 17a-g.
Considering the merging approach used to design the new hybrids,
based on the structure of already known GABABR PAMs, we in-
vestigated if they exhibit/possess/modulate GABAB receptor activity.
The functionality of the new synthesized compounds at these receptors
was studied by using [35S]GTPγS binding assay, a well validated
functional assay for GPCRs, using membranes from rat brain cortex.12
To this purpose, to verify if the compounds act as agonist, antagonists,
GABAB receptor positive allosteric modulator (PAM) or negative allos-
teric modulator (NAM), compounds were tested alone or in the pre-
sence of 10 µM GABA at two different concentrations (2.5 and 25 µM).
The results are reported in Table 1.
Compounds of class A were prepared starting from commercially
available
ethyl
4-chloro-2-(methylthio)pyrimidin-5-carboxylate
(Scheme 1) which was converted to the amino derivative 1 by reaction
MeOH at reflux temperature afforded a 5:2 mixture of the expected
acylhydrazine 2 and the hydrazine derivative 3.
Data are the mean
SEM of 3 experiments, each performed in
triplicate. As the maximal effect of 10 μM GABA alone differed between
experiments, data were normalized to the effect of 10 μM GABA (con-
trol, set as 0%).
Reducing the amount of reagent or changing the reaction solvent
did not improve the course of reaction in favour of compound 2.
Acetylation of 2 to give the acylhydrazide 4 was accomplished by re-
action with acetic anhydride in 18 h. Attempts to reduce the reaction
time by the use of catalytic amount of DMAP resulted in the formation
of the diacetylated compound in 16% yield. Cyclization of 4 to give the
1,3,4-oxadiazole derivative 5 was carried out using POCl3 at 110 °C.
Compound 3, obtained as a side product, was further elaborated in the
corresponding pyrazole 6 by reaction with acetylacetone in MeOH,
using HCl as a catalyst. Finally, 1,2,4-oxadiazole 7 was produced by
alkaline hydrolysis of ester 1, followed by cyclization of acid 8, in the
presence of acetamidoxime, in turn prepared by reaction between
acetonitrile and hydroxylamine, in DMF at 100 °C.
Under our experimental conditions, GABA at 10 µM stimulated
[
35S]GTPγS binding to approximately 22
1.3% of the basal activity
and the GABA-induced activation was antagonized by the orthosteric
GABAB receptor antagonist CGP54626 (data not shown).
Compounds belonging to class A, C and D reduced the basal
[
35S]GTPγS binding to rat cortical membranes in a concentration-de-
pendent manner, while compounds of class B were almost ineffective.
Specifically, at the highest concentration tested 1, 5, 6, 7 (Class A),
13c (Class C), 17b, 17c, 17d and 17f (Class D) induced a decrease of
basal [35S]GTPγS ranging from −50 till −30% respect to basal values.
The other compounds at the highest concentration tested (25 µM) de-
creased less than 15%-20% basal GTPγS binding. Likewise, compounds
6, 7, 13c, 17b, 17c, 17d and 17f in a concentration-dependent manner
significantly decreased the GABA-stimulated [35S]GTPγS binding. All
other compounds only slightly decrease the GABA-induced GTPγS sti-
mulation, thereby resulting substantially inactive.
Compounds of class B were obtained starting from benzylamine
hydrochloride, which was converted to 9 by treatment with benzoni-
trile in the presence of triethylaluminum (Scheme 2). Cyclization to 10a
yethanol and irradiating with microwaves at 150 °C. Repeating the
same synthetic procedure using methyl diethyl malonate and isopropyl
diethyl malonate, resulted in an unsatisfying yield of 2% of 10b and
10c, respectively.
Though initially unexpected, these results are of particular interest
since they might open the avenue to a new class of antagonists/po-
tential negative allosteric modulators of the receptor. In addition, they
confirm that chemical tractability for GABABR PAMs, and more in
general, for GPCR allosteric modulators, cannot be taken for granted.
Indeed, very often, also slight structural modifications of the parent
compound lead to inactive analogues or to functional switches, which
make SAR considerations virtually impossible.23
Nevertheless, further attempts to improve reactions yields, such as
reagent change to the more reactive diphenyl malonate or the use of
DBU coupled with ionic liquid, did not afford the expected results.
OH
O
N
OH
R
R
N
R3
R1
S
R1
R3
O
N
N
H3CS
N
NH
N
O
N
R2
R2
A
B
C
D
Fig. 2. Structural classes of the new compounds.
2